An unfinished page and the question it leaves

The page holds a fugue in four voices, written out in open score so that each line can be followed on its own staff, and the writing proceeds with the patience of a man demonstrating something rather than inventing it. Voice answers voice in strict imitation. A countersubject locks against the subject at every entry. Then, partway through a passage where a third subject has entered and the texture has tightened to its densest point, the hand stops. The bar is complete. The next bar never begins. There is no flourish, no double bar, no sign that the writer believed the argument had reached its end. He was in his middle sixties when those pages were prepared for the engraver, working without a commission, without a performance in prospect, and without any employer who had asked for the work or would ever hear it. He simply stopped, and the most systematic piece of counterpoint he ever attempted has remained an open question ever since.

That page belongs to the Art of Fugue, and the composer was Johann Sebastian Bach, but the names matter less at the outset than the condition the page reveals. Here is the central fact about Bach greatest works as a whole, and the thesis the sections that follow defend from the evidence of the catalogue itself. The output divides into music written to a deadline and music written to no deadline at all. The first kind filled the churches of Leipzig week after week, taught pupils, honoured patrons and paid a salary. The second kind, concentrated in the final decade of his life, answered to nobody. No congregation waited for it, no prince had ordered it, and in several cases no performance took place during his lifetime or for long afterward. Yet it is the undeadlined music, the Mass in B minor assembled for no service, the Musical Offering grown from a single royal theme, and the Art of Fugue left open on its final page, that changed what later composers believed a work of music could attempt. The deadline music proves the craft. The undeadlined music reveals the ambition.

The distinction needs care, because it is not a distinction between lesser and greater quality, and the article returns to that caution repeatedly. A cantata written for a single Sunday in Leipzig could contain a chorus of a complexity that no other composer in Europe would have risked with the forces available, and many of the weekly works repay a lifetime of listening. Nor did the late speculative works arrive from a different mind. The same habits run through both groups: the reuse of strong material, the testing of one idea against every key and every combination, the refusal to publish a solution until its consequences had been worked through. What changes across the catalogue is the pressure under which the habits operated. Deadline pressure produced breadth, hundreds of works fitted to feasts, funerals, lessons and civic ceremonies. Freedom from deadline produced concentration, a small group of vast structures in which a single musical idea is pursued until its possibilities are mapped.

A reader meeting the catalogue for the first time faces a problem of scale before any problem of taste. More than a thousand works carry numbers in the modern catalogue, and they span every genre his century recognised: church cantatas and Passions, concertos and orchestral suites, sonatas for single unaccompanied instruments, teaching pieces for children, and organ music for services he played himself. The sections ahead take the catalogue in groups, work by work, asking of each the same questions. What is the piece, under what circumstances was it written, what does it do that the neighbouring works do not, how was it received, and where should a newcomer place it in a listening order? Biography is compressed throughout, because the life has its own article elsewhere in this cluster, and the theory behind the technique is likewise treated in depth by a sibling study. What belongs here is the output itself, the largest surviving body of work by any composer of the Baroque era, and the strange shape it makes when deadline and freedom are set side by side.

The catalogue and how it is numbered

Order had to be imposed on that output after the fact, because the composer himself published only a small fraction of it and left the rest in manuscript, scattered among his children, his pupils and the institutions that had employed him. The modern numbering comes from the Bach-Werke-Verzeichnis, the catalogue compiled by Wolfgang Schmieder and first published in 1950, which assigns each work a BWV number. The system groups works by genre rather than by date. Cantatas occupy the opening numbers, motets and Passions follow, then masses and oratorios, then chamber and orchestral music, keyboard works, and organ works toward the end of the sequence. A BWV number therefore tells a listener what kind of piece is under discussion, and it tells nothing at all about when the piece was written. A low number is not an early work and a high number is not a late one, a point that confuses newcomers and that the catalogue’s own logic explains once the genre principle is grasped.

How many works did Bach write?

The BWV catalogue assigns numbers to more than 1,100 works and fragments, counting revisions, alternative versions and pieces of doubtful authorship held in a separate appendix of the catalogue. The honest answer carries a range rather than a single figure, because lost works, duplicate versions and disputed attributions move the total at its edges.

What is a BWV number?

A BWV number is the reference assigned to a work in the Bach-Werke-Verzeichnis, the thematic catalogue of the output prepared by Wolfgang Schmieder and published in 1950. The numbers run by genre, so the cantatas stand first and the organ works stand late, and the number identifies a work without claiming a date for it.

Christoph Wolff, whose documentary biography Johann Sebastian Bach: The Learned Musician appeared in 2000, treats the catalogue problem as a problem of survival no less than a problem of counting. A large share of the manuscripts passed at the composer’s death to his sons, and the portions inherited by Wilhelm Friedemann Bach were partly dispersed and partly lost, while the share held by Carl Philipp Emanuel Bach survived in far better order. What the BWV sequence describes, on Wolff’s reading, is therefore not the output as it left the desk but the output as two centuries of keeping, selling, copying and discarding happened to preserve. The gaps are part of the evidence. Whole genres are represented unevenly: the secular cantatas survive in smaller proportion than the sacred ones, the chamber music from the Cöthen years is thinner than the Leipzig church music, and the concertos for several harpsichords hint at a larger lost family of orchestral works behind them.

The principal groups of the catalogue, with the span each group covers and the demand each makes on a first hearing, can be set out together.

Work	Approximate date	Occasion or commission	What it does that no other work in the catalogue does	Difficulty for a first-time listener
Brandenburg Concertos, BWV 1046 to 1051	1718 to 1721	Presentation set for Margrave Christian Ludwig of Brandenburg	Gives each of six concertos a different solo grouping, so no two works share an instrumental argument	Low
St John Passion, BWV 245	1724	Good Friday service at Leipzig	Compresses the Passion narrative into a fast, stark choral drama driven by short choruses	Moderate
St Matthew Passion, BWV 244	1727	Good Friday service at Leipzig	Splits choir and orchestra into two answering bodies and frames the narrative with contemplative arias	Moderate
Christmas Oratorio, BWV 248	1734 to 1735	Six feast days across Christmas and New Year at Leipzig	Rebuilds secular celebration cantatas into a sacred cycle that keeps their festive character	Low
Mass in B minor, BWV 232	1733 to 1749	Assembled without a commission or a liturgical home	Combines movements from three decades into a complete Latin mass of symphonic scale	Moderate
Goldberg Variations, BWV 988	1741	Published as a keyboard exercise volume	Builds thirty variations over the harmony of a single aria rather than over its melody	Low
Cello Suites, BWV 1007 to 1012	around 1720	Written at Cöthen, dedicatee unknown	Sustains full harmony and counterpoint on one four-stringed instrument without accompaniment	Low
Well-Tempered Clavier, BWV 846 to 893	1722 and around 1742	Teaching collections for pupils and family	Works through every major and minor key in prelude and fugue pairs, twice, a generation apart	Moderate
Musical Offering, BWV 1079	1747	Grown from a theme given by Frederick the Great at Potsdam	Turns one royal theme into canons, ricercars and a trio sonata as a single demonstration	High
Art of Fugue, BWV 1080	around 1742 to 1750	Written without commission or performance in prospect	Pursues one subject through a complete architecture of fugues and canons and stops unfinished	High

Two further features of the numbering deserve a place here because they shape how every later section cites the works. The first is the appendix of doubtful and spurious pieces, the Anhang, which holds works that travelled under the famous name without secure evidence behind them. The existence of that appendix is a standing reminder that fame attracts attribution, and one famous case, the Toccata and Fugue in D minor, is examined at length in its own section because the argument there is conducted by named scholars on documentary grounds rather than settled by habit. The second feature is revision. Several works exist in more than one authorised version, the St John Passion among them, and the catalogue handles the versions under a single number with lettered subdivisions. A listener comparing recordings may therefore meet noticeably different music under the same BWV reference, and the difference is a fact about the composer’s working habits rather than an error in the edition.

Deadline music and undeadlined music

The working week explains the first group. From 1723 the composer held the post of Thomaskantor in Leipzig, responsible for music in the city’s principal churches, and the post carried a production obligation that no modern commission schedule resembles. A cantata was required for most Sundays and feast days, with chorales fitted to the readings, rehearsals fitted around a school timetable, and parts copied by hand for each performance. Alongside the church obligation sat the teaching, the civic ceremonies, the collegium musicum concerts he directed from 1729, and the steady traffic of pupils whose lessons generated keyboard pieces as a by-product. Music written inside that system was functional in the strictest sense. It had a date, a place, a set of available performers and a text fixed by the church calendar, and it was usually heard once by the congregation for whom it was made.

Against that system stands the second group, and recognising it is the first rung of this article’s argument. Most of the catalogue was written to a weekly deadline, which changes how its consistency should be read. Consistency under deadline is an achievement of method: a composer who must deliver every Sunday develops procedures that hold under pressure, and the even quality of the Leipzig cantatas is evidence of a system working, not of inspiration arriving on schedule. The late works invert the conditions. The Goldberg Variations were prepared for publication as a considered statement about variation form. The Musical Offering took a theme supplied by a king and answered it with a printed demonstration of contrapuntal possibility. The Mass in B minor gathered movements across sixteen years into a structure no single service could contain. The Art of Fugue pursued one subject through an entire architecture of fugues and canons with no performance arranged at all. Each of these works was written because its composer chose the problem, and each was finished, or left unfinished, on his own timetable.

The biographical frame for that working life, the posts at Weimar and Cöthen that preceded Leipzig and the family circumstances inside which the music was produced, belongs to the companion account of the life and career, which carries the chronology in full. What matters for the works is the residue the frame left in the music itself. Weimar gave the organ works and the first cantata experiments. Cöthen, where the court was Calvinist and church music was not required, gave the instrumental chamber music, the suites for solo strings and the Brandenburg set. Leipzig gave the Passions, the cantata cycles and, in the final decade, the freedom to withdraw from weekly production and write the speculative works. Genre follows employment with unusual fidelity in this catalogue, which is why a works article can be organised, as this one is, by the conditions of production rather than by chronology alone.

One qualification protects the argument from its own neatness. The boundary between the groups is porous, and the composer treated it as porous. Movements written for a Sunday crossed into the Mass. A secular birthday cantata crossed into the Christmas Oratorio. A concerto movement crossed into a cantata sinfonia. The traffic runs in both directions and across the whole career, and a later section takes self-borrowing as its subject in full. The deadline and undeadlined groups are therefore best read as poles rather than as boxes. At one pole stands music whose form was set by an occasion outside the composer. At the other stands music whose only occasion was the problem it set itself. The catalogue’s peculiar greatness lies in the traffic between the poles, in a mind that filed every solution it found, under whatever pressure, against the day a harder problem might need it.

The weekly cantatas as a body of work

A cantata, in the Leipzig of the 1720s, was the principal musical item of the Lutheran service, a setting of poetic and biblical text for solo singers, choir and a small orchestra, lasting roughly twenty to thirty minutes and placed around the sermon. The new Thomaskantor arrived in 1723 and set about supplying the cycle the office demanded at a pace that remains the most concentrated feat of production in the catalogue. Across his first years in the post he composed complete annual cycles, a cantata for nearly every Sunday and feast of the church year, so that the congregation heard new music of his making on most weeks of the year. The second cycle, begun in 1724, was built systematically on chorale melodies, each cantata taking a hymn as its frame and opening with a large chorale chorus. The plan gives that cycle a unity no other group of the works possesses, since the same hymn tune governs the opening chorus and the closing chorale of each piece while the interior movements paraphrase its stanzas.

The forces were modest and the conditions were plain. Leipzig provided a small town band, a choir drawn from the Thomas School whose older boys carried the demanding parts, and soloists who were usually those same pupils rather than hired professionals. The writing takes account of those limits at every turn. Oboes double the strings or carry an obbligato line against a single voice. Trumpets and drums appear on feasts, when the council released the town trumpeters for church duty. The chorus writing is dense but the choral movements are kept short, and the long spans of a cantata belong to the solo arias and recitatives, where one singer and a handful of players could sustain the argument. Reading the scores against the payroll explains features that look like choices and were in part accommodations, although the accommodations were then turned, characteristically, into strengths: the obbligato aria, one voice in dialogue with one instrument, became a signature form precisely because the personnel suited it.

As a body of work the cantatas resist the anthology treatment they often receive. Individual numbers circulate detached from their settings, yet the unit the composer planned was the whole cantata, and behind it the whole cycle, fitted to a calendar of readings that repeats each year. A cantata for a given Sunday answers the appointed Gospel and Epistle, comments on them through its poetry, and closes with a chorale stanza the congregation knew. Heard in sequence across a cycle, the works map a theological year as much as a musical one. That is why the strongest modern defences of the cantatas, among them the complete traversal undertaken by John Eliot Gardiner with the Monteverdi Choir across the year 2000, have presented them in liturgical order, performing each work on or near the feast for which it was written and letting the calendar supply the programme.

One question hangs over the whole body of work, and the article leaves it open here on purpose. The cycles as planned should contain more cantatas than the catalogue holds. Chorales and libretti survive for works whose music does not, pupils’ copies break off, and the composer’s own obituary speaks of five complete annual cycles. What happened to the cantatas that are not in the catalogue is taken up in the section that follows, because the answer changes the status of everything that survives: a sample, however rich, is read differently from a complete set.

How many cantatas survive and how many are lost

The shareable arithmetic of the catalogue can be stated in one sentence. The obituary written by his son Carl Philipp Emanuel Bach with Johann Friedrich Agricola, published in 1754, credits the composer with five annual cycles of church cantatas. Five cycles at roughly sixty works a cycle gives a figure in the region of three hundred sacred cantatas composed. Against that figure, about two hundred sacred cantatas survive complete in the catalogue, with a further small number preserved in part. Nearly one hundred cantatas may therefore be lost, and the surviving body, vast as it is, represents perhaps two thirds of what the Leipzig congregations heard. The numbers are given here as ranges and estimates because the evidence supports nothing firmer, and any account that states a precise loss is claiming more than the record holds.

That estimate is the second rung of the article’s argument. A large fraction of the sacred cantatas is missing, so the surviving body is a sample, not the whole. The sample is not random, and its biases matter for every judgement made about the works. Survival favoured the cycles that remained in Leipzig in performing parts, the works the composer’s widow Anna Magdalena could sell or place, and the manuscripts that passed into careful hands. It disfavoured the earliest Weimar cantatas in some cases, the secular cantatas written for single occasions, and any work whose parts were lent, worn out in performance or divided among heirs. Christoph Wolff’s documentary work traces several of these routes of loss, including the dispersal of material inherited by the eldest son, Wilhelm Friedemann, whose portion of the estate suffered most. The consequence for listeners is plain: the cantatas that dominate concert programmes are the survivors twice over, preserved first by chance and then selected again by editors and performers, and the lost third of the output stands behind every generalisation about the whole.

The evidence for the missing works is concrete rather than speculative. Printed libretti survive for Leipzig cantatas whose music has vanished, since the texts were published for the congregation in annual booklets. Picander, the poet who supplied many of the Leipzig texts, published a complete annual cycle of cantata poems in 1728, announcing that the settings had been composed, yet a portion of that cycle has no surviving music. Individual movements turn up transplanted inside later works, which is how the traffic described in the section on self-borrowing also serves as an archive of the lost: a chorus preserved inside the Mass or an oratorio may be the sole remnant of a cantata otherwise gone. Copyists’ manuscripts occasionally preserve a single aria from a work whose remaining movements disappeared. Each such fragment is counted in the catalogue’s reckoning, and each one confirms that the loss was of complete, performed works rather than of sketches or abandoned drafts.

For the newcomer the practical meaning is reassuring rather than melancholy. Two hundred surviving cantatas still exceed what most listeners will work through in years of listening, and the survivors include the cycles the composer himself valued highly enough to revise and repeat. The loss matters most to the argument of this article, because it disciplines the claims made for the catalogue. When the sections ahead call the cantata cycles the largest body of deadline music in the output, the claim rests on the surviving two thirds. The whole, had it survived, would make the same point at greater length, and the undeadlined late works would stand against an even larger mountain of weekly production than the one that can be heard.

The cantatas worth hearing first

Choice among two hundred works needs a principle, and the principle that serves a first hearing best is contrast of occasion. Begin with a cantata whose scale announces itself in the first chorus. Wachet auf, BWV 140, written for a November Sunday in 1731 and built on a well known hymn, sets its chorale melody in long notes at the centre of a flowing orchestral texture, then answers it with solo movements of unusual directness, including dialogues between two voices that carry the poetry’s image of a bridegroom and a waiting soul. The work is late within the Leipzig output, composed after the great cycles, and it shows the chorale cantata idea refined to a point where every element, hymn, chorus, aria and recitative, sits in exact proportion. Few single works give a clearer first account of what the form could hold.

A second doorway is the solo cantata, where one voice carries an entire argument. Ich habe genug, BWV 82, for bass voice with oboe and strings, written for the feast of the Purification in 1727, contemplates an old man’s readiness to leave the world, and its central aria rocks in a slow triple metre that performers across centuries have treated as a test of expressive control. Jauchzet Gott in allen Landen, BWV 51, stands at the opposite pole: a soprano and a solo trumpet in brilliant dialogue, written for a September Sunday around 1730, demanding a singer who can match the trumpet’s athletic figures phrase for phrase. Between those poles sits Herz und Mund und Tat und Leben, BWV 147, whose closing chorale, a flowing setting in triple time, circulates widely in arrangements detached from the cantata it completes. Hearing the movement in place, after the arias and recitatives that prepare it, restores its function as a conclusion rather than a keepsake.

A third doorway is occasion itself. The Coffee Cantata, BWV 211, is a secular miniature about a father’s attempt to cure his daughter of coffee drinking, performed by the collegium musicum rather than in church, and it shows the deadline composer writing comedy with the same care he gave the liturgy. The Peasant Cantata, BWV 212, written in 1742 for a village celebration, folds folk like dance tunes into its numbers with deliberate roughness. Neither work is representative of the cycles, and that is their use: they mark the range of the cantata habit, from the Thomaskirche on a feast day to a tavern entertainment, and they prepare a listener for the Passions by showing how naturally dramatic pacing came to a composer whose theatre was the church. The deeper account of how such effects are built, the harmony and counterpoint underneath them, belongs to the companion study of the musical style explained, which takes the technique apart in its own terms.

The St John Passion

The St John Passion, BWV 245, was the first large scale work of the Leipzig years, performed on Good Friday in 1724 at the Nikolaikirche, within a year of the composer taking up his post. Its text draws the narrative from the Gospel of John, carried by a tenor Evangelist in recitative, with the words of individual figures given to solo singers and the words of crowds given to the chorus. Around that narrative spine the libretto sets contemplative arias and chorales, in which the action pauses and the event is considered from the standpoint of the listening congregation. The whole spans about two hours in performance, divided into two parts that originally framed the Good Friday sermon. Its scale was a declaration of intent from a new cantor to a city council that had appointed him as a third choice and watched his ambitions with suspicion.

The distinguishing character of the St John setting is speed and edge. The crowd choruses are short, rhythmically driven and harmonically restless, cutting into the narrative rather than decorating it, so that the trial scenes accumulate a pressure that the longer St Matthew setting deliberately avoids. The opening chorus, an orchestral span of churning semiquavers beneath sustained choral cries, establishes the work’s governing tension between agitation and lament before a word of narrative has been sung. The arias are fewer and more concentrated than in the later Passion, and several of them carry an obbligato colour that fixes their place in the memory: a soprano aria with flutes, a bass aria answered by the full chorus, a tenor aria pressed forward by agitated strings. Where the work needs stillness it finds it in the chorales, plain four part settings placed at the joints of the drama.

The work also documents the composer’s revising habit across his whole Leipzig tenure. He returned to the score in 1725 with substantial changes, replacing the opening and closing choruses and several arias, restored much of the first version in a later revival, and was preparing a further revision near the end of his life. The catalogue therefore holds the St John Passion in layered versions, and modern performances choose among them. John Eliot Gardiner, whose performances of the Passions with the Monteverdi Choir were gathered across decades of touring and recording, has written and spoken about the St John setting as the more theatrical of the two, a work whose dramatic compression suits a single evening and whose crowd writing Gardiner treats as the sharpest in the catalogue. That characterisation, from a conductor who has performed both works hundreds of times, is worth holding against the habit of treating the St John as a rehearsal for its successor. It was written first, but it was not written as a draft.

Reception followed the pattern of the whole sacred output. The work was performed in Leipzig across the composer’s lifetime and then fell silent with the rest of the church music, its manuscripts divided among heirs. Its return to public performance came in the wake of the nineteenth century revival of the larger works, and its standing since has been that of the tauter, darker companion: shorter, harder in outline, and often preferred by listeners who find the St Matthew setting’s contemplative breadth slower to enter. For a first Passion, the St John offers the narrative at its most direct, and its demands on performers, particularly on the Evangelist who must sustain the story almost without rest, explain why strong performances of it carry a particular charge.

The St Matthew Passion

The St Matthew Passion, BWV 244, followed in 1727, performed on Good Friday at the Thomaskirche, and it doubles nearly every resource of its predecessor. Where the St John setting uses one choir and one orchestra, the St Matthew divides both into two answering bodies, so that dialogue becomes the work’s structural principle: two choruses question and answer each other, two orchestras interlock, and solo voices are distributed across the divided stage. The libretto, prepared with Picander, draws the narrative from the Gospel of Matthew and surrounds it with reflective poetry on a scale the earlier work does not attempt. The performance span approaches three hours. Nothing in the Leipzig calendar required a work of this size, and the council that paid for the copying of its parts was funding an ambition well beyond the terms of the post that produced it.

The architecture repays description because the scale is carried by structure rather than by volume. The opening chorus builds a procession in sound: two choirs in dialogue above an orchestral texture in a slow triple metre, with a third body of voices, originally boys placed apart from the main choirs, singing a chorale melody above the whole span. The narrative then proceeds through recitative of unusual expressive precision, the Evangelist’s line and the words assigned to Christ surrounded by a halo of strings that falls silent at the final scene. Arias are placed as acts of contemplation after each turn of the story, many of them with obbligato instruments that comment alongside the voice: a viola da gamba, a pair of flutes, an oboe whose line answers the singer phrase by phrase. Chorales recur as pillars, each one harmonised afresh, so that the congregation’s own hymnody is heard commenting in changing harmonic light as the drama darkens.

The work’s afterlife is the best documented case of the catalogue’s disappearance and return. Performed in Leipzig in the composer’s lifetime, in versions he adjusted across revivals into the 1740s, it then left public hearing for generations while keyboard players and theorists studied other parts of the output in manuscript. Its performance in Berlin in 1829 under Felix Mendelssohn, with a large choir and a shortened, reorchestrated score suited to that occasion, is the event from which the public revival of the sacred works is conventionally dated. The Berlin performance did not restore the work as written; it presented an adaptation, and the recovery of the original scoring and scale occupied editors and performers for the rest of the century. The distinction matters to this article’s argument. What returned in 1829 was a monument. What the manuscripts held was a working church piece, written for specific Leipzig forces, and the distance between those two objects is the distance the whole catalogue travelled between deadline and posterity.

Within the catalogue the St Matthew Passion stands as the largest deadline work, the point at which music written for a service reached a scale its service could barely contain. Its defenders across two centuries have usually made the same claim in different words: that the contemplative frame and the narrative are held in exact balance, so that neither the drama nor the meditation subordinates the other. Gardiner, in his writings on the Passions, places the weight differently, stressing the work’s character as a communal act, music designed for a congregation that knew the chorales and heard its own voice inside the structure. Both readings can be held together, and the work sustains them, which is one measure of why it, rather than any other single composition, is the one by which the catalogue is most often introduced to the wider public.

What the two Passions do differently

Set side by side, the two Passions divide the possibilities of the form between them. The St John setting is the more urgent. Its narrative moves quickly, its crowd choruses strike in short hammered phrases, and its arias are placed sparingly, so the drama feels driven from within the Gospel text itself. The St Matthew setting is the more spacious. Its double choir layout turns the telling into a public meditation, its arias are more numerous and more inward, and its chorales speak for a community that is placed inside the events rather than watching them. The difference begins in the source texts. John’s Gospel presents a figure in command of events, and the music keeps a corresponding tautness. Matthew’s account is longer, more detailed and more populated, and the setting expands to hold the detail.

Scoring follows the same division. In the St John Passion a single orchestra of strings, oboes and continuo, with flutes added in the version history, supports one chorus that must be crowd, commentator and congregation in turn. In the St Matthew Passion the divided forces allow functions to be separated in space: one choir can carry the narrative crowd while the other answers as believers, and soloists emerge from within each body. The celebrated dialogues of the later work, in which voices question each other across the two orchestras, are therefore not ornament but method. The earlier work achieves its effects by concentration, the later by disposition, and a listener who knows both can hear the composer’s mind testing the same dramatic problem under two opposite solutions within three years.

The two works also differ in what they ask of their central singers. The Evangelist of the St John must drive a fast moving narrative with fewer resting places, and the tenor arias carry an agitation that matches the crowd writing around them. The St Matthew spreads its solo writing wider, giving its most inward arias to alto and soprano voices framed by single obbligato instruments, and reserving its bass arias for moments of public weight. Conductors who programme both works across a single season, as Gardiner’s Monteverdi Choir has done in its touring, tend to report the same practical contrast: the St John exhausts by intensity, the St Matthew by span. Neither is a preliminary study for the other, and the catalogue is the richer for holding both solutions rather than a single settled manner.

A final difference belongs to reception. The St Matthew Passion became the emblem of the revival and for generations stood for the whole sacred output in the public mind, a position its scale justifies and its history explains. The St John Passion recovered more slowly and has been revalued upward by performers who prize its compression. The honest summary is that the pair should be heard as companion answers, written close together by a composer still in his first Leipzig years, to the question of how a Gospel narrative can be carried in music for a congregation that already knows how the story ends.

The Christmas Oratorio and its reused material

The Christmas Oratorio, BWV 248, is a cycle of six cantatas written for the feast days running from Christmas Day 1734 to the Epiphany of 1735, each part designed for its own service and each framed by the Gospel narrative for its day. Heard complete, the cycle follows the Nativity story from the birth through the shepherds, the naming, the journey of the Magi and the return, held together by the Evangelist’s recitative and by chorales that recur as points of rest. The festive scoring is the fullest the Leipzig works deploy: trumpets and drums open the cycle in bright D major choruses, oboes in their pastoral registers colour the shepherd music of the second part, whose opening sinfonia sets the scene before any voice enters. The work was planned as a cycle from the start, its six title pages describing one oratorio distributed across the season, and its unity is a matter of design rather than of later assembly.

The feature that draws scholarly attention, and that connects the oratorio to the central argument of this article, is its material. A substantial share of the oratorio’s choruses and arias was refitted from earlier secular cantatas, congratulatory works written for the Saxon royal family in 1733 and 1734, whose occasions passed and whose music would otherwise have been heard once and shelved. The opening chorus of the oratorio began as the opening chorus of a birthday cantata. Whole arias changed their texts and their expressive meaning while keeping their notes. The technique, called parody in the scholarship of the period, was standard practice across the Baroque era, and the oratorio is its most prominent monument in the catalogue: a sacred cycle of the first rank built significantly from music composed for court celebrations the year before.

The refitting was skilled work rather than clerical labour. New texts were written to match the existing vocal lines, their phrase lengths, their word painting opportunities and their placing of emphasis, so that the poetry of the oratorio was in part composed to fit music that already existed. Where the original movement’s character resisted the new subject, it was not used. The result complicates any simple account of expression in the works, since a chorus that sounds to a listener like the inevitable musical image of rejoicing shepherds began as rejoicing at a royal birthday. Christoph Wolff, in his studies of the composer’s working methods, treats the oratorio’s construction as evidence of an economy of invention: strong material was capital, held against future need, and the composer’s judgement lay in knowing which material could bear which new text. The next section takes that economy as its whole subject.

Reception has treated the oratorio kindly and partially. Its six parts are most often performed singly or in selections during the Christmas season, and the festive choruses circulate as seasonal fixtures. Heard whole across its six days, or in a single complete performance, the cycle reveals the architecture that the seasonal excerpts conceal: the return of the opening chorale melody at the end, the symmetry of the trumpet led outer parts around the quieter centre, and the progressive darkening as the story moves toward Herod’s threat. Among the large sacred works it is the most immediately approachable, its choruses direct, its dance rhythms open, and its borrowed brilliance no less effective for having been brilliance first spent elsewhere.

Self-borrowing as normal practice

Reusing his own material was standard practice and not a lapse, which complicates any argument about inspiration. That is the third rung of this article’s argument, and the works supply the evidence in every genre. Concerto movements became cantata sinfonias and choruses. The opening chorus of one cantata reappears as a movement of the Mass. A hunting cantata’s aria becomes an oratorio aria. Keyboard pieces were transcribed for other instruments, organ works for orchestra, and whole concertos reshaped for harpsichord from versions for violin or oboe that in some cases survive and in others must be inferred from the refitting itself. The practice spans the career from the Weimar years to the final decade, and it was conducted openly, in a musical culture that treated a composer’s existing music as a stock to be drawn on rather than as a set of fixed objects each tied to a single occasion.

Three reasons explain the habit, and all three are practical before they are aesthetic. The first is the deadline system itself: a composer owing a cantata a week, with copying and rehearsal inside the same week, held his earlier material as insurance, and a movement that had succeeded in one setting could be trusted in another. The second is occasion. Secular cantatas for birthdays and civic ceremonies were performed once, for patrons who would not hear them again, and their music was too valuable to retire after a single hearing; transferring it into church works or published collections gave it a durable home. The third is quality control. The movements chosen for reuse were disproportionately the strong ones, and the refitting usually involved revision, rescoring and adjustment rather than plain copying. Self-borrowing in this catalogue is therefore evidence of judgement exercised twice: once in the writing and once in the selection of what deserved to travel.

The practice also disciplines the language listeners use about the works. A Romantic account of composition, in which each piece issues from a unique impulse fitted to its subject alone, cannot survive contact with the oratorio’s borrowed choruses or the Mass’s assembled movements. Richard Taruskin, in the Oxford History of Western Music published in 2005, presses the wider point that eighteenth century music making was a craft economy of reusable material, and that the modern cult of the unique, inviolable work was projected backward onto composers who did not share it. Taruskin’s characterisation is of the period’s practice rather than of one man’s expedients, and it fits the evidence here exactly. What distinguishes this catalogue is not the borrowing, which was universal, but the standard of the material borrowed and the intelligence of its placement. A borrowed chorus in the Mass in B minor sits inside an architecture planned around it, and detecting the join requires the documents rather than the ear.

For performers the traffic between works has a practical consequence worth stating. A movement known in one setting teaches the other. Singers who have met a melody as an oratorio aria approach its cantata original with the phrasing already in the voice, and editors reconstruct lost originals from surviving refittings with a confidence the documents support. The lost cantatas of the earlier discussion are partly recoverable by exactly this route, since movements preserved inside later works carry the fingerprints of their first forms. Self-borrowing, the habit that looks like a shortcut, turns out to be one of the catalogue’s preservation mechanisms, and the works are best understood as a single connected stock of material that the composer spent a lifetime arranging, rearranging and refining.

The Mass in B minor

The Mass in B minor, BWV 232, is the catalogue’s largest single structure and its most puzzling commission, in that there was no commission. Across the years 1733 to 1749 the composer assembled a complete setting of the Latin Mass Ordinary, the Kyrie, Gloria, Credo, Sanctus and Agnus Dei, drawing some movements from earlier works, composing others new, and revising the whole into a span of roughly two hours for voices and orchestra on a scale beyond any service he ever directed. No complete performance took place in his lifetime. No Lutheran service could have contained it, since the Leipzig liturgy used only portions of the Latin Ordinary and never the complete cycle at this length. The work therefore stands as the supreme case of the undeadlined group: a mass assembled across sixteen years, for no patron, no feast and no congregation, by a composer in his sixties who cannot have expected to hear it whole.

Its history begins with a document that was meant to be usable. In 1733 the composer sent the Kyrie and Gloria, a Missa in the terminology of the time, to the Elector of Saxony in Dresden, accompanied by a petition for a court title. The parts were prepared with unusual care, each one written out for the Dresden performers, and the dedication succeeded in its worldly purpose: the title of court composer followed in 1736. Those two movements, however, became only the opening of the later structure. Around 1748 and 1749, in the final years, the composer returned to the score, added a Credo built partly from refitted cantata movements and partly from new writing, completed the Sanctus from a setting first heard in Leipzig in the 1720s, and closed the work with an Agnus Dei and a final chorus that reuses the music of an earlier thanksgiving cantata movement in transformed guise. The architecture was planned at the end, around material gathered across a career.

The scale of individual movements explains the work’s standing among performers. The Kyrie opens with a choral span of a gravity that announces symphonic intent. The Gloria deploys trumpets and drums in festive D major and then withdraws to duets of chamber intimacy. The Credo sets its central articles in music of deliberate archaism, plainchant quoted in long notes beneath modern counterpoint, as if the age of the text were itself being represented. The Sanctus, for six voice parts, builds a choral texture of a density the Passions never require. Singers who have performed the work describe a physical experience unlike any other in the catalogue: the choral writing sustains its intensity across spans that test breath, tuning and concentration together. That testimony, repeated by choral conductors across generations, is part of the evidence for the work’s position, since the difficulty is structural rather than incidental and was plainly accepted by its composer as the price of the design.

The question the work poses can be left open here in the form in which the composer left it. Why would a Lutheran composer write a full Latin mass he could never use? The next section takes the question directly, because the answers proposed by scholars go to the heart of how the late works should be understood: as professional strategy, as confessional statement, as a summary assembled for posterity, or as some compound of the three.

Why a Lutheran wrote a Latin mass

The puzzle dissolves in part when the confessional facts are set out plainly. Lutheran Leipzig retained Latin music inside its services to a degree that surprises readers who picture a clean break with the older church. Latin settings of the Kyrie and Gloria were sung on feast days, Latin motets formed part of the repertory, and the Latin school attached to the Thomaskirche kept the language in daily use. A Lutheran cantor composing Latin movements was therefore working inside his own tradition, not borrowing a foreign one. The 1733 Missa sent to Dresden went further in practical terms: Dresden was a Catholic court, and a Latin mass setting was the appropriate currency for a petition addressed to its Elector. The opening movements of the great structure thus had both a liturgical home in part and a diplomatic purpose in whole.

Scale is where the puzzle returns, and it is the completed Credo, Sanctus and Agnus Dei of the final years that require explanation, since no petition was pending then and no Dresden performance was arranged. The strongest account offered by Christoph Wolff reads the completed mass as a deliberate summary: a composer aware of age, assembling his most considered sacred writing into the most comprehensive form his tradition offered, and ordering it as a legacy. On that reading the refitted cantata movements inside the Credo are not admissions of haste but selections, the best sacred choruses of the Leipzig decades gathered into a permanent frame. The work becomes a catalogue within the catalogue, a retrospective exhibition mounted by its own maker, and its lack of a performance becomes intelligible: the audience it addresses is whoever keeps the score.

A second account stresses craft rather than testament. The mass form offered the one large structure in which every sacred manner the composer commanded, archaic counterpoint, concerted splendour, intimate duet, fugal choruses of several kinds, could be displayed in sequence under a single text whose authority no patron needed to approve. On that reading the work is a demonstration, like the Musical Offering and the Art of Fugue, of what the art could contain, addressed to connoisseurs and to the future rather than to a court. The two accounts differ in emphasis more than in substance, and the documents support elements of each: the care of the autograph score, the absence of performing parts prepared for any occasion, and the timing of the completion in the final years all point the same way. What would settle the matter, a letter stating the intention, does not survive, and the honest conclusion is structural: the mass was completable only by a composer free of deadline, and it was completed. Reception supplied the performance history the composer did not, with the complete work reaching a full public hearing in Leipzig in 1859, more than a century after its assembly, by which time its scale read as prophecy rather than as surplus.

The motets

The motets occupy a small and distinct corner of the catalogue: a handful of works for unaccompanied choir, most of them written for funerals and memorial occasions in Leipzig, where the genre retained a function that the cantata had elsewhere displaced. Six motets are securely attributed, with a small number of further candidates at the edges of the canon. Their scoring is for double choir in the largest cases, voices alone without independent instruments, the continuo sometimes supporting and sometimes absent. In a catalogue dominated by concerted music, the motets stand apart by their means: everything must be done by voices in counterpoint, and the writing accepts that discipline at its strictest.

Singet dem Herrn ein neues Lied, BWV 225, for two four part choirs, is the most brilliant of the group, its outer sections built on antiphonal exchanges that throw phrases between the choirs at gathering speed, its centre a chorale treated in long notes against a free counterpoint in the second choir. Jesu, meine Freude, BWV 227, is the most architectural: eleven movements arranged symmetrically around a central fugue, alternating chorale stanzas in varied harmonisations with movements setting a biblical passage, so that two texts interleave across the whole span. The funeral motets written for particular Leipzig citizens, among them Der Geist hilft, BWV 226, composed for the burial of the Thomas School rector in 1729, show the genre at work in its actual setting, where the choir that sang it stood in a documented relation to the person commemorated.

The motets carry a reception history disproportionate to their number. They were among the sacred works that remained in use after the composer’s death, since the Thomaskirche choir kept motet singing in its practice and the manuscripts stayed where they had been used. Mozart, visiting Leipzig in 1789, heard Singet dem Herrn performed by the school choir and, on the account transmitted by Rochlitz, asked to study the score, an episode that belongs to the works’ transmission whether or not every detail of the report is exact. For listeners the motets offer the clearest available evidence of the choral counterpoint in isolation, unassisted by orchestra, and choirs treat them as examinations: intonation, stamina and balance are exposed in every bar. Their modest count should not mislead. Pound for pound of performing forces, they are among the most concentrated music in the catalogue.

The Brandenburg Concertos

The six Brandenburg Concertos, BWV 1046 to 1051, were presented in 1721 to Margrave Christian Ludwig of Brandenburg in a manuscript of exceptional calligraphic care, accompanied by a dedication in French of elaborate courtesy. The works themselves were not new for the occasion. They were gathered from the Cöthen years, when the composer directed the court orchestra of Prince Leopold and had a body of skilled instrumentalists at his daily disposal, and several of them refit earlier material. What the presentation created was the set: six concertos, each with a different solo grouping, arranged as a demonstration of every way a concerto could be constituted. No two works in the set share an instrumental argument, and that variety is the set’s thesis.

The groupings repay enumeration in prose because each one defines its work. The first concerto fields a festive body of horns, oboes and strings with a small solo violin. The second sets a high trumpet, recorder, oboe and violin as four soloists of wildly unequal power and balances them with a sleight that performers still treat as a benchmark. The third is for strings alone, three violins, three violas and three cellos in interlocking parts over continuo. The fourth gives its solo writing to a violin and two recorders. The fifth makes the harpsichord a soloist for the first time in the history of the form, building to a cadenza of a length that turns the accompanist’s instrument into the protagonist, a part the composer plainly wrote for his own hands. The sixth removes the violins altogether and gives the lead to violas, with cellos and gambas below them, producing the darkest sonority of the set. A listener who works through the six in order hears the concerto idea dismantled and rebuilt six times, and no other group in the catalogue covers its genre so completely.

The concertos also mark a stylistic meeting point. Their ritornello structures and solo episodes follow the Italian concerto model associated with Vivaldi, whose works the composer had transcribed for keyboard in Weimar, while their inner part writing keeps the density of German counterpoint. David Ledbetter, in his writings on the keyboard works and the period’s practices, situates such meetings as the normal condition of the era’s music: forms travelled, and a composer’s character shows in what the imported form is made to carry. In the Brandenburg set the imported form is made to carry a systematic survey of instrumentation itself, which is why the set reads, beyond its brilliance movement by movement, as a treatise that happens to be delightful. The circumstances of its delivery, and the silence that followed, belong to the next section, because they contain one of the catalogue’s most instructive ironies.

Why the Brandenburg set was probably never played for its dedicatee

The dedication was, in substance, a job application. The composer had met the Margrave in Berlin in 1719, while collecting a new harpsichord for the Cöthen court, and the Margrave had asked to see some of his work. The presentation manuscript followed two years later, with a dedication that recalled the request, praised the Margrave’s taste and held the composer’s services at his disposal. The timing aligns with trouble at Cöthen, where Prince Leopold’s marriage had cooled his interest in his orchestra and the establishment faced cuts. A Berlin post, or Berlin favour, was worth courting, and the six concertos were the portfolio of a Kapellmeister presenting the full range of what he could write for an orchestra. That is the fourth rung of this article’s argument: the Brandenburg set was a job application that appears never to have been performed by the man it was sent to.

The evidence for the silence is documentary. The Margrave’s Berlin establishment was small, a handful of musicians whose number and skills fall well short of the forces the concertos require, and several of the works demand specialists, a horn pair, a high trumpet, a harpsichord soloist of the first rank, that the Berlin chapel did not employ. No parts for performance were copied from the presentation score, which was a fair copy for the eye rather than a working set for stands. The manuscript entered the Margrave’s library, was valued at his death in 1734 in an inventory at a nominal sum among bundles of concertos by many hands, and passed in time to other collections. No payment, no appointment and no letter of thanks from Berlin survives. The application failed, and the composer stayed at Cöthen until the Leipzig post opened in 1723.

The irony needs no heightening. The set that failed to move its dedicatee became, after the nineteenth century recovery of the manuscripts and the first complete publication of the concertos in the Bach Gesellschaft edition, the most widely known orchestral music in the catalogue, recorded in dozens of cycles and used wherever a single work must represent the Baroque concerto. The presentation score itself, preserved through the library descent described in the documents, is the principal source for the works, which means the portfolio that was never played is the reason the music survives in authoritative form. A listener deciding between the two morals on offer, that quality guarantees its audience or that survival is a chain of accidents, will find the Brandenburg history supporting the second at every step. The works are in the catalogue because a clerk filed a manuscript, not because a prince applauded.

The orchestral suites

The four orchestral suites, BWV 1066 to 1069, gather the French dance suite into orchestral dress: an overture in the French manner, its slow dotted opening answered by a fast fugal section, followed by a sequence of dances and character pieces. The works belong in part to Cöthen and in part to Leipzig, where the collegium musicum concerts gave orchestral music a public home outside the court. Their scoring grows across the set. The first two suites rest on oboes, bassoon and strings in the French taste. The third and fourth add trumpets and drums, and with them a civic splendour suited to the Leipzig performances, in which the suites likely honoured public occasions whose exact record has not survived.

The third suite, in D major, BWV 1068, contains the movement by which the set is known to the widest public: the slow movement for strings that circulates in an arrangement for violin and piano under a title referring to the string on which its melody can be played. The arrangement, made in the nineteenth century by the violinist August Wilhelmj, transposed and simplified its source, and the distance between the arrangement and the movement in its place, second in a five movement suite and scored for the full string body, measures the difference between the catalogue as excerpted and the catalogue as written. In context the movement is a moment of suspension inside a festive structure, and its effect depends on the overture’s brilliance before it and the dances after it. The fourth suite answers it with the fullest ceremonial writing of the set, while the second suite, for a solo flute with strings, holds the most intimate scale and ends with a badinerie whose speed has made it a proving piece for flautists.

As a group the suites are the catalogue’s most social orchestral music, written for listening in company rather than for church or study, and their comparative neglect beside the Brandenburg set says more about programming habits than about quality. A suite asks a concert to accept dance movements as its substance, and the third and fourth works repay the acceptance fully. For the newcomer they offer the easiest entry into the orchestral output: the forms are open, the movements short, the pleasures immediate, and the craft, the part writing inside the dances and the pacing of each overture, reveals itself on repeated hearing without any need for preparation.

The violin concertos

The surviving violin concertos are two works for a single violin, in A minor and E major, BWV 1041 and 1042, and one concerto for two violins in D minor, BWV 1043, all associated with Cöthen or with the Leipzig collegium years. The count is certainly incomplete. The harpsichord concertos of the Leipzig period refit concerto material wholesale, and the internal evidence of their solo writing, its figuration lying naturally under a violinist’s hand and awkwardly under a keyboard player’s, indicates lost originals behind several of them, including a D minor concerto reconstructed by scholars for violin and performed in that form. The surviving two solo concertos and the double concerto are therefore the visible portion of a larger concerto practice, and they should be read as representatives of a genre the composer worked in more extensively than the archive shows.

The A minor concerto is the most concentrated: three movements of economy and edge, its finale a gigue of relentless momentum. The E major concerto is the most expansive, opening with a ritornello of festive confidence and placing at its centre a slow movement of sustained lyricism over a walking bass pattern that repeats while the solo line floats free of it. The double concerto is the most celebrated, and its slow movement is the reason most often given. Two violins weave lines around each other in close imitation above a spare accompaniment, neither leading for long, the harmony carried by their interlacing. The passage has become a standard test of how two soloists listen, since its effect collapses if either player presses forward. The outer movements keep the same equality at speed, the soloists exchanging the argument in exact balance.

The concertos’ reception mirrors the catalogue’s general pattern in miniature. Known to the nineteenth century through the Bach Gesellschaft edition, they entered the repertory as violinists adopted them, and the double concerto in particular became a fixture of teaching and of concert programmes, a work senior players perform alongside students as a rite of passage. Their scale suits them to the role. Each concerto lasts about a quarter of an hour, states its argument without surplus, and leaves a listener with a complete account of the composer’s concerto manner, the ritornello discipline, the slow movement’s patience, the finale’s release. For a newcomer they belong early in any listening order, and their lost companions, recoverable in outline from the harpsichord versions, stand as one more reminder that the catalogue’s surviving shape is partly an accident of paper.

The sonatas and partitas for solo violin

The six works for unaccompanied violin, three sonatas and three partitas, BWV 1001 to 1006, survive in an autograph manuscript dated 1720, written in a hand of calligraphic precision that performers treat as both a musical text and a physical object of beauty. Nothing else in the catalogue asks so much of a single player. A violin has four strings and one bow, and polyphony on it must be implied, broken and suggested more often than it can be sounded outright. These works build fugues, dances and variations inside that constraint, implying bass lines, inner voices and harmonic rhythm from a single melodic strand supplemented by double stops and chords. The sonatas follow the church sonata pattern of slow and fast movements with a fugue at the centre of each. The partitas are dance suites, each gathered around a character of its own.

The achievement is partly one of notation. The composer devised, across these six works, a complete language for suggesting counterpoint on a melodic instrument: arpeggiated chords that distribute a harmony across the strings, bariolage figures that alternate a repeated note against a moving line, and melodic writing whose leaps outline the missing voices so the ear supplies them. Generations of violinists have used the set as the foundation of technique and musicianship together, and the works entered the repertory early in the nineteenth century recovery, championed by the virtuosos who first performed them complete in public. Their difficulty is of a particular kind. The notes lie within a competent player’s reach more often than Romantic virtuoso writing does, but the voicing, the decision about which strand of an implied texture to bring forward at each moment, is left to the player, so two performances of the same movement can differ as readings while agreeing on every pitch.

The set’s architecture also rewards whole hearing. The three fugues grow in scale across the sonatas, the last, in C major, building the longest span. The partitas move between the intimate and the monumental, the first compact, the second closing with the Chaconne that the next section treats alone, the third in E major opening with a preludio of pure moto perpetuo brilliance that the composer himself later refitted as a cantata sinfonia and as an organ work. That migration of a single movement across instruments and functions returns to the article’s recurring theme: material in this catalogue is capital, and the solo violin works, for all their apparent solitude, are connected by working threads to the church and the organ loft.

The Chaconne and its standing

The Chaconne closes the second partita, in D minor, BWV 1004, and it outweighs the movements that precede it: a set of variations over a repeating bass pattern, running about a quarter of an hour in performance, longer than the rest of its partita combined. The form is the most restrictive in the set. A chaconne holds its harmonic pattern fixed and asks invention to renew itself inside the repetition, variation after variation, without modulation to fresh ground for relief. Across its span the writing moves through every resource the solo violin possesses: chordal writing of four part density, rapid figuration, arpeggios that sweep the strings, passages of bare single line, and a central section in the major mode that offers the work’s only sustained release before the minor variations resume and close the argument at its starting gravity.

The standing of the piece among violinists is a documented fact rather than a flourish of this article. Johannes Brahms, writing to Clara Schumann in 1877 after studying the work, described the experience of a composer producing, on one staff for a small instrument, a world of thought and feeling, and Brahms made a left hand piano arrangement of the piece as an act of study. Ferruccio Busoni later arranged it for full piano, and transcriptions for guitar, for orchestra and for other instruments followed across the twentieth century, each one evidence that players of other instruments wanted the argument in their own medium. No other movement in the catalogue has attracted so many acts of appropriation, and the reason performers give is consistent: the Chaconne is complete. It lacks nothing that its single instrument withholds, and the work of the transcribers has been homage rather than repair.

For listeners the Chaconne offers a particular kind of attention test. Its variations do not announce themselves with changes of tempo or character in the manner of a variation set on a theme tune. The ground repeats, and the interest lies in watching figuration, register and density transform above the repetition until the pattern itself feels less like a constraint than like a landscape crossed in changing weather. David Ledbetter’s writings on the period’s variation practices place the piece within a tradition of chaconne composition that includes French and Italian models, and the comparison sharpens rather than dims its standing: the models are accomplished, and none sustains the form at this length or this intensity. It is the single movement most often proposed when the whole catalogue must be represented by one instrumental work, and the proposal survives scrutiny.

The cello suites

The six suites for unaccompanied cello, BWV 1007 to 1012, parallel the violin set and differ from it in almost every quality that matters. Where the violin works press against the limits of polyphony, the cello suites breathe: each suite opens with a prelude and continues through the standard dance sequence of allemande, courante, sarabande and gigue, with a pair of lighter galanteries inserted before the final dance. The writing sustains harmony and implied counterpoint on a single instrument, as the violin works do, but the cello’s register lets the implied bass be a real presence, and the suites’ character is correspondingly grounded, their rhetoric closer to speech and dance than to demonstration. No autograph survives. The principal source is a copy made by Anna Magdalena Bach, the composer’s second wife and a professional copyist in her own right, supplemented by copies from his circle, and the small variations among those sources have occupied editors for generations.

The suites were composed around 1720 at Cöthen, where the court employed an accomplished cellist, and their early history after that is unusually faint even by the catalogue’s standards. They were published in the nineteenth century and used largely as studies, their status as concert music unrecognised, until Pablo Casals, who had found a copy in a Barcelona shop as a boy in the 1890s, studied them for years and brought them into public performance and, in the late 1930s, into recordings that reached a wide public. The Casals account of the suites’ recovery is the catalogue’s clearest case of a performer single handedly changing a work’s category, from exercise to repertory, and it happened within living memory of the recording era’s start. Everything in the suites’ modern standing descends from that advocacy: the complete cycles by cellists that followed, the adoption of the first prelude as the instrument’s most recognised solo opening, and the use of the suites as the measure by which young cellists are judged.

Their content rewards the standing. The preludes alone map six temperaments, from the first suite’s flowing arpeggios to the fifth’s dark scordatura writing, in which one string is retuned to produce resonances unavailable in standard tuning, and the sixth suite’s bright expanse, written for a five stringed instrument whose added high string opens a register the others cannot reach. The sarabandes are the still centres, several of them among the most concentrated slow movements in the catalogue. For a newcomer the cello suites are the most hospitable large structure in the instrumental output, dance based, humanly scaled, and complete in a single sitting per suite, and they are placed early in the listening order proposed near the end of this article for exactly those reasons.

The Well-Tempered Clavier as two books

The Well-Tempered Clavier is two collections, each of twenty four preludes and fugues, one pair for every major and minor key, the first book completed in 1722 and the second assembled around 1742. The title page of the first book states its purpose in terms that define the teaching thread of the catalogue: for the profit and use of musical youth desirous of learning, and for the recreation of those already skilled in the study. Well tempered refers to a tuning system in which every key is playable, as distinct from the older mean tone tunings that made remote keys unusable, and the collections demonstrate the freedom such a tuning grants by actually composing in all twenty four keys, a completeness no earlier keyboard collection had attempted. The point was practical as much as theoretical: a pupil who worked through the books had met every key, every common figuration and a graduated range of fugal procedure.

The two books differ in character across their twenty year interval. The first is the more pedagogical in its sequencing, its preludes testing patterns of figuration one at a time, its fugues mostly compact. The second book is larger in scale and freer in manner, its preludes approaching the character of sonata movements and inventions for two voices, its fugues more spacious, the whole reflecting a composer revisiting his own demonstration with the resources of his maturity. David Ledbetter’s book length study of the first collection, published in 2002, reads the 1722 volume as a designed curriculum rather than an anthology, tracing how its ordering teaches harmony, fingering and composition together, and Ledbetter’s account has shaped how teachers present the work since. The fugues, meanwhile, became the discipline’s central texts: generations of composers learned counterpoint by copying and analysing them, and the practice of calling the collection a testament of keyboard art, whatever words are used, rests on that documented classroom history rather than on piety.

Performance questions surround the books in a way that suits this article’s purposes. The title says Clavier, keyboard, without specifying harpsichord, clavichord or organ, and the works have been played on all of them and on the modern piano, whose sustaining pedal and dynamic range reshape the textures in ways performers defend and dispute. The question of whether the pairs should be heard in sequence is taken up in the FAQ section, since it is one of the decisions a new listener actually faces. What belongs here is the collection’s double identity: a teaching book that became a concert repertory, and a demonstration of a tuning system that became the common practice foundation for all later keyboard music. Few works in any catalogue are simultaneously a method, a manifesto and a body of pieces that survive extraction one by one. These two books are, and their position in the catalogue is correspondingly central.

The inventions and sinfonias as teaching pieces

Below the Well-Tempered Clavier in scale and above it in directness stand the fifteen inventions and fifteen sinfonias, the two and three part contrapuntal pieces composed in the early 1720s for the instruction of the composer’s eldest son, Wilhelm Friedemann, whose keyboard book contains their early versions. The autograph fair copy of 1723 carries a title page that states the method with a teacher’s precision: an honest guide by which lovers of the keyboard, and particularly those desirous of learning, are shown a clear way not only to play cleanly in two parts and then in three, but also to arrive at good inventions and to develop them properly, and above all to acquire a singing manner in playing. Every element of that programme matters. Invention, in the period’s usage, means finding a musical idea and working it, and the pieces are lessons in composition disguised as lessons in performance.

Their construction is a model of economy. Each piece takes a single short motif and builds its whole span from it, by imitation between the voices, by sequence, by inversion and by exchange of parts, so that a pupil who masters a piece has also mastered the procedure that generated it. The two part inventions move from the clearest cases, imitation at the octave in a home key, toward chromatic and remote key territory. The three part sinfonias add the management of a middle voice, the skill on which all later contrapuntal playing depends. Teachers have kept the set in continuous use for three centuries on the strength of that design, and the pieces survive the classroom: played complete, in sequence, they form a recital span of genuine variety, since the motifs range from dance like figures to severe chromatic subjects.

The teaching thread in the catalogue deserves emphasis because it corrects the picture of a composer writing either for church or for posterity. A large body of the keyboard output was written for named pupils and family members, at the desk, for hands he knew. The notebook collections assembled for Anna Magdalena, with their minuets, chorales and small pieces by several hands including hers, belong to the same domestic economy. The inventions stand at the head of that thread because their purpose is written on their title page and their method is audible in every bar. Where the Well-Tempered Clavier demonstrates what the keyboard can contain, the inventions teach how a player begins to contain it, and the two collections together explain why the composer’s reputation among musicians, as distinct from audiences, was secured in the classroom long before the concert hall recovered the Passions.

The English and French suites

The two named sets of keyboard suites carry their national labels from later tradition rather than from their composer, and the labels point in misleading directions. The six English Suites, BWV 806 to 811, composed in the Weimar and Cöthen years, are the grander set, each opening with a substantial prelude and following the standard dance sequence at spacious length. The six French Suites, BWV 812 to 817, from the early Leipzig years, are smaller in scale, without preludes, their dances compact and their manner intimate. Neither set has a secure connection with its named country beyond style associations perceived by early biographers: the French suites are indeed lighter in texture, in a manner the period called French, while the English set’s grandeur may reflect a dedication or a copyist’s title now lost to view. The names stuck, and the catalogue keeps them as a warning about how labels attach to works for reasons the works themselves do not supply.

Their musical substance needs no label. The English Suites contain some of the most ambitious keyboard writing of the composer’s earlier years, their preludes approaching concerto scale, their gigues fugal and driven, their sarabandes carrying ornamented repeats that show the written out embellishment practice of the period in full. The French Suites, by contrast, have been the entry point for generations of students, their dances proportioned for developing hands while their part writing stays exact. Heard as a pair of sets, they display the suite form’s range in the hands of a composer who treated the dance sequence not as a relic but as a frame that could hold both display and intimacy. Alongside them stand the partitas for keyboard, published one by one from 1726 and collected as the first part of the Clavier-Übung in 1731, the composer’s first published works by his own choice, which enlarge the suite to its greatest scale in the catalogue and include movements, a French overture among them in the second partita’s companion volume, that reach beyond the dance frame entirely.

Publication matters in this corner of the catalogue. Most of the output circulated in manuscript and reached print late or never, but the keyboard works were the exception the composer himself managed. The partitas were engraved at his expense and sold by subscription and through agents in several cities, with the title page advertising the composer by name and office. The suites copied for pupils spread hand to hand. Between those two channels, manuscript for the household and print for the public, the keyboard music built the reputation that survived the eighteenth century when the church music had fallen silent, and the suites, published and copied alike, were its travelling representatives.

The Goldberg Variations

The Goldberg Variations, BWV 988, were published in 1741 as the fourth part of the Clavier-Übung, the series of keyboard collections the composer issued in print across his later years, and the title page describes the work with a modesty that borders on misdirection: an aria with diverse variations, for a harpsichord with two manuals, composed for music lovers for the recreation of their spirits. The aria is a slow sarabande of poised and ornate character. Thirty variations follow. The whole spans well over an hour with its repeats observed, and it is the largest keyboard work in the catalogue, a structure whose scale is matched by the completeness of its plan. Publication was an act of intent. Unlike the church music left in manuscript, this work was engraved, priced and sold, addressed to connoisseurs the composer would never meet.

The early account of the work’s origin is famous and must be handled with the caution the brief for this article requires. Johann Nikolaus Forkel, the composer’s first biographer, writing in 1802, reported that the variations were commissioned by Count Hermann Carl von Keyserlingk, the Russian ambassador in Dresden, for his young harpsichordist Johann Gottlieb Goldberg, to be played during the Count’s sleepless nights. The story gives the work its common name and its favourite anecdote. It is not corroborated by the documents. The published title page names no dedicatee, which a commission of that rank would normally have proclaimed, Goldberg was about fourteen when the work appeared, and no payment or letter survives. The responsible treatment, followed here, is to present the tale as Forkel’s report, possible in outline and unproven in detail, and to note that the work needs no anecdote: its design is its occasion. The FAQ section returns to the story and its evidence directly.

What the work indisputably is can be stated from the score. A single aria generates thirty variations organised in groups of three, every third variation a canon, the canons proceeding by widening intervals from the unison to the ninth across the set, while the variations between the canons explore hand crossing brilliance, dance characters, arabesque figuration and, at the centre, a slow variation of a depth that performers treat as the work’s heart. The final variation is a quodlibet, a weaving together of popular tunes in counterpoint, a family joke of a kind the composer’s relatives are documented to have enjoyed, after which the aria returns, unchanged. The return reframes everything heard since its first statement. Few structures in the catalogue close their own circle so completely, and the next section takes the architecture apart in its own terms.

The aria and the architecture above it

The Goldberg design rests on a distinction that explains its inexhaustibility. The variations are not variations on the aria’s melody. They are built over the aria’s bass line and harmonic plan, its chord sequence held constant while everything above it, melody, rhythm, texture, character, is reinvented in each variation. A listener following the tune will lose it within the first variations and may conclude the thread has been dropped. A listener following the harmony hears the same ground plan return thirty times in thirty kinds of light, and the work’s coherence is located there. The method is the chaconne principle of the D minor partita enlarged to the scale of a whole evening, and it connects the Goldberg Variations to the catalogue’s deepest habit: hold one element fixed, and test everything else against it.

The canon ladder gives the architecture its visible spine. Every third variation is a canon, a strict imitation between voices, and the interval of imitation widens by step across the set, from canon at the unison through canon at the second, the third and onward to the ninth, after which the quodlibet takes the place the canon at the tenth would have occupied. The scheme is symmetrical around the central fifteenth variation, the slow canon at the fifth, written in the minor and in an inverted form, which stands as the design’s dark midpoint. Groups between the canons alternate character pieces and virtuoso studies, several of them requiring the two manuals of the specified harpsichord so the hands can cross and layer without collision. On a single manual piano the same passages demand redistribution of notes between the hands, one of the practical questions performers of the work must answer for themselves, and the difference in sonority between the two instruments is part of the work’s modern performance debate.

Reception completed the architecture’s fame. Published for connoisseurs, the variations were admired by specialists across the nineteenth century without entering wide currency, and their modern standing dates from the recording era: Wanda Landowska’s harpsichord recording of the 1930s presented the work as a monument of its own instrument, and Glenn Gould’s piano recording of 1955, made when he was twenty two, presented it as a structure of thought, taken at speeds and with a clarity that divided opinion and fixed the work at the centre of the piano repertory in a single release. Gould recorded it again in 1981 at markedly slower tempos, and the pair of readings, by one player across a lifetime, has become a standard exhibit in any discussion of how interpretation changes while a text stays fixed. For the listener the practical counsel is simple: hear the aria first and last, follow the bass, and let the variations accumulate. The architecture reveals itself by repetition, which was the composer’s method in building it.

The organ works and the chorale preludes

The organ was the composer’s first instrument and his public credential: his appointments turned on organ examinations, his reputation among contemporaries rested at least as much on his playing as on his compositions, and the organ works run from his teens to his final years. The free works, the preludes and fugues, toccatas, fantasias and the great Passacaglia in C minor, BWV 582, belong largely to the Weimar years, when he served a ducal chapel with a fine instrument and composed for his own performance. The Passacaglia builds twenty variations over an eight bar ground bass and then converts the ground into the subject of a fugue, the Goldberg method stated a generation earlier at organ scale. The six trio sonatas, BWV 525 to 530, composed in Leipzig as studies for Wilhelm Friedemann, set two hands on separate manuals against the pedals as three equal voices, chamber music of a transparency that organists treat as the purest test of independence between the limbs.

The chorale preludes, settings of hymn melodies for organ, form the devotional counterpart. The Orgelbüchlein, the little organ book begun in Weimar, planned more than one hundred and sixty chorale settings arranged by the church year and completed fewer than fifty of them, its title page promising a method by which a beginning organist learns to play a chorale in every manner and to gain facility in the pedals. The completed settings are miniatures of extreme concentration, each one fitting its chorale melody inside a texture generated from a single motif, so that a piece of two minutes carries the same unity of invention as the inventions carry at the keyboard. The later chorale collections enlarge the scale: the eighteen large preludes assembled in Leipzig, and the third part of the Clavier-Übung, published in 1739, which frames catechism chorales of two sizes inside a prelude and fugue of public grandeur, a published organ mass in all but name.

Peter Williams, whose multi volume study of the organ music set the scholarly standard for the field across the editions published from 1980 onward, reads the organ output as the composer’s most continuous autobiography in music: the early works absorbing the north German masters he walked to Lübeck to hear as a young man, the Weimar works consolidating a personal manner, the Leipzig publications summarising and teaching. Williams’s characterisation is consistently that of a documentary scholar, weighing sources and registrations before reputations, and his scepticism in the case taken up next is continuous with that method. The organ works also hold the catalogue’s most practical afterlife. Church organists play the chorale preludes in services for which their companions were written, week by week, in an unbroken functional tradition unlike anything else in the output: deadline music that kept its deadline.

The Toccata and Fugue in D minor and its disputed status

The most famous organ piece attributed to him survives in no autograph and may not be his. That is the fifth rung of this article’s argument, and the section that follows presents the dispute as the live scholarship it is, with a named case on each side. The Toccata and Fugue in D minor, BWV 565, is known to a public far beyond organ music: its opening flourishes and its storm of semiquavers have made it the popular emblem of the instrument and of its supposed composer. The work survives in a single manuscript copy by Johannes Ringk, a pupil of a pupil, with no autograph, no early print and no documentary trail reaching the composer’s desk. Attribution rests on that copy’s heading and on tradition. No other major work in the catalogue depends on so narrow a thread.

Did Bach write the Toccata and Fugue in D minor?

The attribution is disputed on documentary and stylistic grounds and remains unsettled. The case for the work rests on its attribution in the surviving copy and on passages consistent with early keyboard writing, while the case against it, argued by Peter Williams, rests on the absence of an autograph and on features without parallel in the authenticated organ works.

The case against, as Peter Williams set it out in an article of 1981 and maintained in his subsequent writings, is cumulative. The opening’s parallel octaves, the fugue’s construction, the passage work’s reliance on patterns the authenticated works use sparingly, and the piece’s overall proportions each sit outside the documented habits of the organ music, and together they suggest either a work by another hand or a transcription: Williams proposed that the piece may have originated as a work for solo violin, its textures translating violin figuration in ways that would explain several of its oddities at the organ. The proposal matters because it converts an argument about quality, which the piece’s fame renders useless, into an argument about medium, which evidence can test. Later scholars have divided. Some defend an early work, pointing to experiments in the authenticated youthful pieces and noting that a young composer may write once in a manner he later abandons. Others extend the doubt toward the wider circle of north German organists and pupils whose works travelled under stronger names.

The stance taken here follows the evidence discipline this series requires. What is settled is the documentary record: one copy, no autograph, a late and indirect attribution. What is disputed is the inference from style, and inferences from style require parallels that scholars weigh differently. What would settle the matter is a source closer to the composer, an autograph, a copy by a family member or a Leipzig pupil, or a contemporary reference to the piece by name, and no such source has appeared. The work therefore stays in the catalogue under its number, played and recorded as it has been for two centuries, with its question mark intact. For the listener the dispute changes nothing in the hearing and everything in the labelling, and the labelling is precisely where honesty is owed: the most recognisable bars in the organ repertory may belong to the catalogue’s appendix in spirit long before any editor moves them there in print.

The Musical Offering

In May 1747 the composer travelled to Potsdam, where his son Carl Philipp Emanuel served as harpsichordist to Frederick the Great, and the king, a flautist and composer of competent ambition, played him a theme of his own invention and challenged him to improvise upon it. The theme is long, chromatic and irregular, plainly designed as a test. The visitor improvised a three part fugue on it at the keyboard that evening and undertook to work the material further. Within months the Musical Offering, BWV 1079, was engraved and sent to the king: a three part ricercar and a six part ricercar built on the royal theme, a set of canons on it of increasing ingenuity, including canons that resolve only when a page is turned or a part read upside down, and a trio sonata for flute, the king’s instrument, violin and continuo, whose movements are likewise grown from the theme. The whole collection is a demonstration addressed to a single reader, that one theme could sustain every contrapuntal procedure the art possessed.

The canons carry inscriptions that frame the gift as both homage and puzzle, and the six part ricercar stands as the collection’s weightiest single movement, written in a learned manner that the composer also preserved in an autograph version. Contemporary report, transmitted through the composer’s family, held that the king’s theme was prized by its recipient less warmly than its compliment deserved, and the collection’s publication turned a court episode into a public document. Like the Goldberg Variations it was printed. Like the mass it was assembled without a performing occasion in view. It belongs with the late group in everything except its origin, which was the most deadline like event imaginable: a king, an evening, a command performance. The distance between that origin and the published artefact, a set of puzzles for connoisseurs to solve at leisure, measures the composer’s habit of converting obligation into speculation.

One further question opens here and is deliberately held for its own section. The Musical Offering and the mass show the late manner at work on given material, a royal theme and a liturgical text. The last great project applied the same manner to material of the composer’s own choosing, a single subject invented for the purpose, pursued through an entire architecture of fugues and canons, and left, at the summit of the design, unfinished in mid phrase. Why the final work stops where it stops is the question the next two sections approach, first by describing the work itself and then by weighing the explanations the documents allow.

The Art of Fugue

The Art of Fugue, BWV 1080, is the work with which this article opened, and its plan can now be stated in full. A single subject, plain and grave in character, generates the entire collection: a sequence of fugues, called contrapuncti in the published edition, that treat the subject in its original form, in inversion, in stretto, in mirror forms where the whole texture can be turned upside down and remain correct, and in combination with new countersubjects, followed by canons at widening intervals. The scoring is deliberately unspecified. The work is written in open score, four staves, in the manner of a theoretical demonstration, and it has been performed by keyboard players, string ensembles, wind groups and orchestras, each medium claiming a different aspect of the argument. The autograph shows the collection growing across the 1740s, and the edition prepared for engraving in the composer’s last years presents it as a summa: everything the fugue, the form he had practised since his teens, could be made to do.

The intellectual character of the work is best grasped through its mirror fugues, movements composed so that the entire counterpoint sounds correctly when inverted, every interval flipping direction while the harmony holds. Writing one such movement is a feat of planning. The collection contains two, in three and four parts. The canons that follow apply the same severity at smaller scale. Nothing in the design addresses a congregation, a court or a classroom of beginners. The address is to the art itself and to the small company of musicians equipped to follow it, and the work’s reception has matched that address: admired by theorists from the start, published after the composer’s death by his son Carl Philipp Emanuel with a notice explaining its unfinished state, allowed to fall out of print, and recovered by the nineteenth and twentieth centuries as the catalogue’s most concentrated monument. Performers who take it up complete do so as a campaign, and recordings of the whole span on a single instrument remain events in a way that recordings of single fugues are not.

The final fugue is the collection’s intended summit and its open wound. It combines three subjects, worked together in a quadruple fugue whose fourth subject was to be the collection’s own main theme, and the third of its subjects spells, in the German notation of the period where the letter B denotes B flat and H denotes B natural, the letters of the composer’s name: B, A, C, H. A composer signing his summit fugue with his own name, in notes, inside a design that was to crown the whole architecture, is the clearest possible evidence that the ending was planned. The manuscript of that fugue breaks off at bar 239, in mid argument, with the three subjects in combination and the fourth still to enter. The section that follows asks why, and the question is the first of this article’s open loops, now due for its answer.

Why the Art of Fugue stops

Why is the Art of Fugue unfinished?

The documents support several explanations and compel none of them. The composer’s health failed in his final two years and his eyesight deteriorated to the point of an unsuccessful operation, while the engraving of the collection was in progress, and the strongest reading is that illness interrupted a work its author intended to complete.

The deathbed version of the story comes from the family. A note in the hand of Carl Philipp Emanuel on the autograph of the final fugue records that the author died while working on the fugue in which the name BACH appears as a countersubject. Forkel’s biography converted that note into a scene, and the scene hardened into legend: the blind composer, dictating or reaching for the pen, stopped by death at the exact bar where his name enters the music. The scene should be resisted in its details. The handwriting of the final fugue’s surviving autograph is firm and belongs, on the palaeographic evidence assembled by Christoph Wolff and others, to an earlier period than the last illness, which suggests the fugue was drafted before the final decline and left among papers awaiting revision or completion of its design. Wolff’s characterisation of the documents, consistent across his writings, is that of a scholar who lets paper outrank pathos, and the paper here points away from a deathbed interruption and toward a project overtaken by the ordinary arithmetic of age, eyesight and competing final works.

The competing explanations deserve their strongest forms. One holds that the work is complete as designed, that a quadruple fugue combining all four subjects was never intended, and that the final fugue’s break is a fragment preserved by the family rather than a ruin of the plan; against that reading stand the son’s own published notice, which calls the work unfinished, and the absence of any closing movement after the canons in the edition. A second holds that leaves of the autograph were lost after the composer’s death, so that a finished ending existed and perished; against it stands the continuity of the surviving manuscript, whose final page ends the fugue in mid bar without a stub or a gap in the gathering. A third, the illness account, fits the known facts of the final years, the failed eye operation of 1750, the months of blindness and decline, and the engraving project left for the family to finish. What would settle the matter is documentary: a completed autograph leaf, a copyist’s full score, or a family letter describing the ending. None exists. The loop opened at the Musical Offering therefore closes with a verdict of weight rather than of certainty: the weight of the evidence places an intended completion interrupted by failing health ahead of its rivals, and the mid phrase stop that opens this article is the trace of that interruption.

The late works as a group

Draw the final decade together and a pattern emerges that no single work states alone. The Goldberg Variations of 1741, the canonic variations on a Christmas chorale, the Musical Offering of 1747, the completed Mass of 1749 and the Art of Fugue across the decade share a set of decisions. Each takes a single given element, an aria’s harmony, a royal theme, a liturgical text, one fugue subject, and builds a complete architecture upon it. Each was prepared for print or for the archive rather than for a dated performance. Each addresses connoisseurs, posterity or the art itself rather than an employer. And each was produced after the weekly obligation had slackened, as the cantor’s performing duties narrowed and the speculative projects of a lifetime moved to the centre of the desk. That is the sixth and final rung of this article’s argument. The late unfinished works were written for no occasion, no patron and no performance, and they are the ones that redefined the ambition of the form.

The redefinition can be specified rather than merely praised. Before these works, a collection demonstrated its completeness by covering its ground, as the Well-Tempered Clavier covers the keys. The late works demonstrate completeness by exhausting a single idea, and the shift from coverage to exhaustion is the shift later composers inherited when they built variation cycles, thematic transformations and whole works from single motifs. The downstream story, who took which procedure and what they made of it, from the contrapuntal studies of Mozart and Beethoven to the fugal finales of the nineteenth century, is traced in the companion account of the influence on music, which owns that territory. What belongs here is the recognition that the undeadlined group is not a retirement hobby appended to a working catalogue. It is the catalogue’s destination, the place where fifty years of deadline solutions were gathered, abstracted and stated as principles.

Richard Taruskin’s treatment of the period supplies the necessary counterweight, and his characterisation of the late works’ reputation is the sceptical one he applies throughout the Oxford History: the elevation of the speculative works above the church music reflects, in part, the values of the scholars and performers who recovered them, readers and analysts for whom written complexity is the natural measure of worth. The counterweight is just, and this article has tried to honour it by treating the cantatas at their full weight. Yet the documents keep their own counsel. The composer himself chose print for the speculative works and left the cantatas in manuscript. He himself spent his final strength on the mass and the fugue collection rather than on a sixth cantata cycle. The late group is the destination because its author walked toward it, and the walk is recorded in paper, not projected onto it.

A listening order for a newcomer

Which Bach piece should I listen to first?

Begin with the third Brandenburg Concerto or the first cello suite, where the invention is immediate and the scale is welcoming, then take the Goldberg Variations for the architecture, one Passion for the drama, and a handful of cantatas across the church year before attempting the Mass or the Art of Fugue.

The reasoning behind that order is a gradient of demand. The instrumental works ask only for attention to sound and pattern, and they give their pleasures quickly: the Brandenburg Concertos for brilliance, the cello suites for intimacy, the double violin concerto for balance. The Goldberg Variations then introduce the catalogue’s long span habit in its friendliest form, one harmony explored for an evening, and a listener who has followed them owns a method for everything larger. The choral works reverse the proper sequence for many guides, and deliberately so here. A Passion heard before the listener knows the sound world can feel like a monument visited out of duty, while the same work heard after the instrumental music arrives as drama in a familiar language. One cantata a week, following the church calendar in a complete cycle recording, remains the deepest route of all, and listeners with the patience for it should take it once the landmarks are known. Readers who keep listening notes across a long traversal of the catalogue may find that an offline encrypted notes app, obtained from the VaultBook download page, suits the purpose of holding a private listening journal that travels between devices on no one’s server but their own.

Two practical counsels complete the order. The first concerns performance style. Period instrument recordings, with their lower pitch, gut strings and smaller choirs, and modern instrument recordings, with their larger forces and sustained sonorities, present the same works in genuinely different lights, and the FAQ section takes up the choice directly, since neither is a substitute for the other and most listeners end by keeping both. The second concerns completeness. This catalogue rewards sampling less than it rewards finishing things: a whole suite, a whole Passion, the whole Goldberg span, the whole Well-Tempered book across a week of sittings. The works were designed as wholes, their proportions are part of their meaning, and the newcomer who finishes six complete works knows the composer better than the sampler who has heard sixty openings. Listeners preparing a talk or a class on the catalogue from their notes can assemble the material in a browser-based presentation viewer such as the ReportMedic presentation viewer, which opens slides without an installed office suite and keeps the session inside the browser.

What the catalogue adds up to

Add the groups together and the catalogue’s shape is a working life made visible. Roughly two thirds of it was produced inside institutions, against calendars, for performers whose names and abilities shaped the writing, and it constitutes the largest body of functional music of the first rank in the European record. The remaining portion, the teaching collections, the published keyboard volumes and the speculative late works, was produced at the composer’s own direction, and it constitutes the period’s most complete statement of what its musical art could contain. Neither portion explains itself without the other. The late works are inconceivable without the decades of weekly solutions behind them, and the weekly works are misread if their composer’s speculative ambitions are forgotten, since the same mind that supplied a Sunday cantata in three days was filing its inventions against the problems of the Art of Fugue.

The reception history, traced across these sections work by work, adds the final term to the sum. The catalogue survived by fragments and accidents, a clerk’s filing of a failed application, a widow’s sale of manuscripts, a school choir’s unbroken motet tradition, a boy in Barcelona buying an edition of the cello suites, and it returned to public life in stages across two centuries, each stage recovering a different portion and mistaking that portion, for a while, for the whole. The frame for that long story within the division is set by the series account of the history of classical composers, where the recovery of this catalogue stands as the central case of a canon rebuilt from paper. The comparative question of where its author stands among composers altogether is owned by the series survey of the greatest composers of all time, and this article does not compete with it. Its own claim is narrower and, on the evidence assembled, defensible: the catalogue divides into music written to a deadline and music written to no deadline at all, and the late works that nobody commissioned are the ones that changed what music could attempt. The unfinished page with which the article began is therefore not an emblem of failure. It is the catalogue’s most honest self portrait, a mind stopped by mortality in the middle of an argument it had every intention of finishing, surrounded by the largest completed body of work its century produced.

Frequently Asked Questions

Q: What are Bach’s most famous works?

The works by which the widest public knows the catalogue are the Brandenburg Concertos, the Goldberg Variations, the two Passions, the Mass in B minor, the cello suites and the opening prelude of the first book of the Well-Tempered Clavier. The Toccata and Fugue in D minor, BWV 565, belongs on any honest list of famous pieces heard under the name, with the caution that its attribution is disputed by scholars on documentary grounds. Fame and scale do not coincide neatly here. The cantatas, which form the largest body of the output, are famous among musicians as a group while individual cantatas reach a smaller public, and the Art of Fugue is famous as an idea, the unfinished summa, among listeners who have not yet heard it complete. A balanced introduction samples each kind.

Q: What are the Brandenburg Concertos?

The Brandenburg Concertos are six concertos, BWV 1046 to 1051, gathered from the Cöthen years and presented in a calligraphic manuscript to Margrave Christian Ludwig of Brandenburg in 1721. Each concerto constitutes the form differently. One is for festive horns, oboes and strings, one balances a high trumpet against recorder, oboe and violin, one is for strings alone in nine interlocking parts, one gives the solo lines to violin and recorders, one makes the harpsichord a soloist with an extended cadenza, and one removes the violins and leads with violas. Together the six works form a survey of what a concerto could be, which is why the set, rather than any single member, is the unit performers record and listeners cite. No two share an instrumental argument.

Q: How were the Brandenburg Concertos commissioned?

They were not commissioned in the contractual sense, and the distinction explains their history. The composer met the Margrave in Berlin in 1719 while collecting a harpsichord for the Cöthen court, and the Margrave asked to see some of his music. Two years later the presentation manuscript arrived, six concertos refitted or gathered from works written for Prince Leopold’s orchestra at Cöthen, prefaced by a dedication of elaborate courtesy that held the composer’s services at the Margrave’s disposal. In substance the set was a portfolio supporting an application, sent at a moment when the Cöthen establishment faced cuts and a Berlin post was worth seeking. No payment, appointment or recorded performance at the Margrave’s court followed, and his chapel lacked the specialist players several concertos require. The manuscript was filed in his library.

Q: What is the difference between the two Passions?

The St John Passion of 1724 is the more compact and urgent setting, built for one choir and one orchestra, driven by short and restless crowd choruses, with fewer arias and a faster narrative pace. The St Matthew Passion of 1727 divides choir and orchestra into two answering bodies, runs close to three hours, and surrounds its narrative with a larger frame of contemplative arias and chorales, so that meditation and drama hold equal weight. The source Gospels shape the contrast: John’s account presents a figure in command of events, while Matthew’s is longer and more populated. Performers who know both describe the St John as exhausting by intensity and the St Matthew as exhausting by span. Neither work is a draft for the other; they are companion solutions to one dramatic problem.

Q: Why did Bach reuse his own music?

Reuse was the normal economy of a composer working under deadline in a culture that treated existing music as capital. A Leipzig cantor owing a cantata most weeks, with copying and rehearsal inside the same week, held strong earlier movements as insurance. Secular cantatas written for a single birthday or civic occasion would otherwise have been heard once and retired, so transferring their best movements into church works, oratorios and the Mass gave them a durable home. The refitting was skilled labour rather than copying: new texts were written to match existing vocal lines, scoring was adjusted, and weak material was passed over. The opening chorus of the Christmas Oratorio began in a royal birthday cantata, and movements of the Mass in B minor began in cantatas. Selection, in this practice, was a second act of judgement.

Q: What is the Mass in B minor and when was it performed?

The Mass in B minor, BWV 232, is a complete setting of the Latin Mass Ordinary, Kyrie, Gloria, Credo, Sanctus and Agnus Dei, assembled between 1733 and 1749 for voices and orchestra at a span of roughly two hours. Its opening Kyrie and Gloria were sent to the Elector of Saxony in Dresden in 1733 with a petition for a court title. The remaining movements were completed in the final years, partly from refitted cantata movements and partly from new writing. No Lutheran service could contain the whole work and no complete performance took place in the composer’s lifetime. Portions circulated in the nineteenth century as editors recovered the score, and the first complete public performance is documented in Leipzig in 1859, more than a century after the assembly was finished.

Q: What are the Goldberg Variations?

The Goldberg Variations, BWV 988, published in 1741 as the fourth part of the Clavier-Übung, are an aria followed by thirty variations for a two manual harpsichord. The variations are built not over the aria’s melody but over its bass line and harmonic plan, which returns unchanged while melody, rhythm and texture are reinvented above it. Every third variation is a canon, and the canons widen by step from the unison to the ninth across the set, giving the design a visible spine. The penultimate variation is a quodlibet weaving popular tunes in counterpoint, after which the aria returns exactly as first heard. The whole work spans well over an hour with repeats observed, and it is the largest keyboard composition in the catalogue, addressed on its title page to music lovers for the recreation of their spirits.

Q: Is the Goldberg insomnia story true?

The story comes from Johann Nikolaus Forkel, the composer’s first biographer, writing in 1802, more than sixty years after the work appeared. Forkel reported that Count Hermann Carl von Keyserlingk commissioned the variations for his young harpsichordist Johann Gottlieb Goldberg, to be played in an adjoining room during the Count’s sleepless nights. No document corroborates the account. The published title page names no dedicatee, which a commission of that rank would normally have proclaimed, Goldberg was about fourteen at publication, and no payment, letter or court record survives. The tale is possible in outline, since Keyserlingk knew the composer and Goldberg was a gifted pupil in his circle, and it gave the work its common name. Scholarship treats it as an unproven biographical tradition rather than as a fact of the work’s origin.

Q: What are the cello suites?

The cello suites are six suites for unaccompanied cello, BWV 1007 to 1012, composed around 1720 at Cöthen. Each suite opens with a prelude and follows the dance sequence of allemande, courante, sarabande and gigue, with a pair of lighter movements inserted before the gigue. The writing sustains harmony and implied counterpoint on one instrument, the cello’s register allowing the implied bass to register as a real presence. No autograph survives; the leading source is a copy by Anna Magdalena Bach. The suites served mainly as studies until Pablo Casals, who had studied them from boyhood, brought them into concert life and recorded them in the late 1930s. The fifth suite retunes one string for darker resonance, and the sixth was written for a five stringed instrument with an added high string.

Q: Why is the Chaconne considered so demanding?

The Chaconne closes the second partita for solo violin, BWV 1004, and sustains variations over a repeating bass pattern for about a quarter of an hour, longer than the rest of its partita combined. Its demands are of three kinds. The harmonic ground never changes, so invention must renew itself through figuration, register and density without relief from fresh harmony. The violin must imply full counterpoint on four strings, sounding chords, broken harmonies and single lines in a sequence that keeps several voices present in the listener’s ear. And the span itself tests concentration, since the performer must pace an architecture whose midpoint, a section in the major mode, is its only sustained release. Brahms studied the work and arranged it for piano left hand, one documented measure of the regard in which players hold it.

Q: How many cantatas are lost?

The estimate rests on the obituary published in 1754 by Carl Philipp Emanuel Bach and Johann Friedrich Agricola, which credits the composer with five complete annual cycles of church cantatas. Five cycles at roughly sixty works each give a figure in the region of three hundred sacred cantatas composed. About two hundred survive complete in the catalogue, with a few more preserved in part, so nearly one hundred cantatas may be lost. The figure is an estimate and should be quoted as one: printed libretti survive for some missing works, a text cycle published by Picander in 1728 lacks music for part of its span, and movements preserved inside later works show that complete performed cantatas vanished. Loss fell unevenly, hardest on material dispersed among the heirs.

Q: What is the Art of Fugue trying to do?

The Art of Fugue, BWV 1080, takes a single fugue subject and submits it to a complete architecture of procedures: plain statements, inversion, stretto, augmentation, mirror fugues whose whole texture remains correct when turned upside down, fugues combining the subject with new countersubjects, and canons at widening intervals. Written in open score without specified instruments, the collection behaves as a demonstration of what the fugue form can contain when one idea is pursued to exhaustion. Its final fugue was designed to combine three subjects with the main theme as a fourth, and its third subject spells the composer’s name in the German pitch notation of the period. The work stops in mid argument at bar 239 of that fugue. Read whole, it is less a set of pieces than a proof, conducted in sound, about the resources of counterpoint.

Q: Should I start with the choral or the instrumental works?

Start with the instrumental works, because they ask only for attention to sound and pattern and they repay it at once: a Brandenburg Concerto for brilliance, a cello suite for intimacy, the double violin concerto for balance between two soloists. Follow with the Goldberg Variations, which teach the long span habit in a friendly form, one harmony explored across an evening. Then take the choral music in rising scale, a single cantata, the Christmas Oratorio, one Passion, and the Mass in B minor when the sound world is familiar. The choral works heard too early can feel like monuments visited from duty, since their texts, chorales and liturgical frames presume acquaintance. Heard in sequence after the instrumental music, the same works arrive as drama in a known language. The one exception is the listener who sings: join a choir and start wherever the choir starts.

Q: Is the Well-Tempered Clavier meant to be played in order?

The two books were designed as collections that proceed through all twenty four major and minor keys, and their title page of 1722 addresses them to musical youth desirous of learning and to the recreation of those already skilled. Playing or hearing a book in sequence is a legitimate and instructive route, since the key order is systematic and the difficulty broadens as the pairs proceed. It is not the only intended use. The preludes and fugues were built to be extracted, studied individually and matched to a pupil’s stage, and most historical use of the books has been selective in exactly that way. Modern complete recordings invite whole traversal, and the second book, assembled around 1742, rewards it with a larger scale than the first. A practical course is to learn the first book selectively and hear the second book whole.

Q: How do I work through the Goldberg Variations as a listener?

Give the work three hearings with three different assignments. At the first, follow the aria’s bass line and its harmonic plan, and listen for its return beneath every variation, since the variations are built over that ground rather than over the aria’s tune. At the second, track the canons: every third variation is a strict imitation, and the interval widens from the unison to the ninth, so counting the groups of three reveals the architecture while the ear enjoys the character pieces between them. At the third, listen for the midpoint, the slow minor variation at the centre of the set, and for the quodlibet near the end, where popular tunes are woven in counterpoint before the aria returns unchanged. Observe the repeats on at least one hearing; the proportions were designed with them in place.

Q: How do I choose between period and modern instrument recordings?

Choose by the quality that matters most for the work in hand, and expect to keep both kinds. Period instrument performances use lower pitch, gut strings, smaller choirs and, for keyboard works, the harpsichord or clavichord, which clarifies counterpoint and lightens textures; the Passions and cantatas gain narrative speed and textual bite from that clarity. Modern instrument performances bring larger forces, sustained string tone and the piano’s dynamic range, which can give the Mass a grandeur and the Goldberg Variations a singing line that period practice achieves by other means. Neither approach is a reconstruction of a single original sound, since the works were heard in varied conditions in their own time. The reliable method is comparative: take one work in two readings, note what each reveals, and let the comparison, rather than a doctrine, build the collection.

Q: Which five Bach works give the best overview?

Five works can map the catalogue if each stands for a different condition of its making. The third Brandenburg Concerto stands for the Cöthen instrumental years and the concerto manner at its most concentrated. The cantata Wachet auf, BWV 140, stands for the Leipzig church output and the chorale cantata in its refined late form. The St Matthew Passion stands for the deadline works at their largest scale. The Goldberg Variations stand for the published keyboard music and the variation architecture built over a single harmony. The Mass in B minor stands for the late undeadlined group, assembled across sixteen years for no service and no patron. Heard in that order, the five works carry a listener from court to church to study, and from music written for a Sunday to music written for posterity, across the whole argument of this catalogue.

Q: What should I listen for in a cantata?

Listen first for the frame: most cantatas open with a chorus or an instrumental sinfonia that states the work’s character, and close with a plain four part chorale, the congregation’s hymn, harmonised afresh. Between the frame posts, follow the alternation of recitative, where the text is narrated or argued in a speech like line over continuo, and aria, where one idea is contemplated at length by a solo voice, often in dialogue with a single obbligato instrument whose line comments beside the voice. Note the chorale melody when it returns in different dress, since in the chorale cantatas one hymn governs the whole design. Note also the scoring’s economy: oboes doubling strings, trumpets reserved for feasts, a gamba or flute appearing for one aria. The text’s appointed reading for the Sunday explains the imagery, and a translation repays keeping open.

Q: How did reception of Bach’s works change over time?

Reception moved in stages, each recovering a different portion of the output. In the decades after 1750 the church music fell from use while keyboard players and theorists studied the fugues in manuscript, so the reputation survived among musicians before it returned to audiences. The Berlin performance of the St Matthew Passion in 1829 under Felix Mendelssohn opened the public revival of the sacred works, in an adapted and shortened form that later generations of editors replaced with the original scoring. The Bach Gesellschaft, founded in 1850, published the complete surviving output across fifty years and made the whole catalogue available to scholarship. The recording era then widened each work’s public in turn, from Casals and the cello suites to Gould and the Goldberg Variations, so that by the end of the twentieth century the catalogue stood as central repertory in every genre it contains.

Q: Why does Bach transfer to other instruments so well?

The music transfers because its interest is carried by counterpoint and harmonic design rather than by the sonority of one instrument. A fugue subject, a bass line and their combinations remain intelligible when the timbre changes, since the argument lives in the relation of the voices. The composer modelled the practice himself, refitting violin music for keyboard, keyboard music for orchestra, and a solo violin prelude as a cantata sinfonia and an organ work, treating his own textures as portable. Later musicians followed the licence he had demonstrated: Brahms and Busoni arranged the Chaconne for piano, the cello suites are played on viola, bass and guitar, and the Art of Fugue, written in open score with no instrument specified, has been recorded by string ensembles, wind groups and orchestras. Transfer succeeds here because the works were conceived as structures first and sonorities second.