It seems likely that Bach began work on his second Passion early in 1725, perhaps for performance on Good Friday that year. He had performed Johannes-Passion in 1724 and, in the end, had to repeat it in 1725 with some changes from the earlier performance; he also had to resort to performing Reinhard Keiser’s St Mark Passion in 1726. However, the long labours eventually bore fruit. This performance is based on a new edition of Bach’s first version of Matthäus-Passion which was probably first performed on Good Friday 1727.

Until the issue of the 2004 Bärenreiter publication edited by Andreas Glöckner our access to this early version (BWV244b) was limited to a 1740s manuscript made by Johann Christoph Altnickol, a student and son-in-law of the composer and to an incomplete score copied sometime after 1741 by J F Agricola, another Bach pupil. There is also a libretto by Picander, the pseudonym used by Bach’s librettist Christian Freidrich Henrici, which was published in 1729. There are, however, clear errors in the Altnickol manuscript which may well also be unfinished. All these sources informed the edition used for this performance. We can only presume that Altnickol’s manuscript is a copy of Bach’s own manuscript version—though it is unclear why it was made when Bach had himself produced a beautifully laid-out manuscript of a revised version, which provides copious performing information, when he had more time to do so in 1736. There is no doubt that Bach’s later version includes new information with, for example, a considerable number of basso continuo figures and some wonderfully vivid composer’s ornamentation—for example, in the soprano aria ‘Blute nur’ where Bach’s decoration of the vocal line on the words 'ermorden' and 'schlanger' brilliantly clarifies the meaning and emotion of the words which is, of course, the primary purpose of ornamentation.

Altnickol’s manuscript is almost devoid of bass figures and instrumental articulations and the resultant implied bowing-patterns and wind tonguings; Bach’s original performing parts may have had more information or some may have been added during rehearsal. The same may well be true of the copious essential ornaments—cadential trills, appoggiaturas in recitative, arias and choruses, ornamental signs and so on. In such instances, the later version may well inform us in indicating what additional stylistic matters Bach expected of his performers in 1727 alongside what they would have expected to do anyway; stylistic expectations of bowing and tonguing patterns were well established by 1727. Similarly, the various ornamentation conventions were also in place as we can see from Bach’s own teaching and manuscripts; these would have included the essential ornaments mentioned above but also the composer’s ornamentation, some of which would have been in the manuscript and some of which Bach or his performers may have suggested in rehearsal. We can’t know what performers’ ornamentation was added though it is not impossible that some of this was approved and admired by Bach and incorporated into his later manuscript. Various stylistic conventions and the resultant articulations and rhythmic alterations would have been familiar or would have become familiar to Bach’s performers.

An interesting element of this version is that it shows Bach using ‘false’ ornamentation in recitatives in the same manner that he had employed in the earlier Johannes-Passion and which he retained for the later version of Matthäus-Passion. This involves indicating an appoggiatura either by an ornamental sign or in standard proportional notation for a masculine (single syllable) phrase end. The employment of an appoggiatura at the end of phrases varied from composer to composer but it seems quite clear that Bach followed the convention of not expecting an appoggiatura at masculine phrase ends but giving the performers the option of doing so for feminine (strong-weak) endings. This means that on the occasions where he indicates one at a masculine ending then it sent a visual signal to the performer and an aural one to the listener. An example of this practice occurs in number 38c with Peter’s third denial of being a ember of Jesus’ followers with the text 'Ich kenne des Menschen nicht' ('I do not know the man!'). The ‘false’ ornament draws attention to Peter’s false statement. The same device is used also in Johannes-Passion. A similarly important moment occurs at the end of number 47 where Pilate’s final syllable (which wouldn’t normally expect an appoggiatura on 'Was hat er denn Übels getan?' ('Why, what evil has he done?') and certainly not one from above for a question) introduces and draws our attention to the content of the subsequent text: 'Er hat uns allen wohlgetan … Sonst hat mein Jesus nichts getan … Aus Liebe will mein Heiland sterben' ('He has done good to us all … Apart from that, my Jesus has done nothing … My Saviour is willing to die for love.') Bach, both by his ornamentation and very special (and surprising) orchestration clarifies the central message of this Passion.

For this recording we have presumed Bach would have encouraged and expected these and other conventions and we have employed them accordingly. A specific instance worth mentioning is with the appoggiaturas in the vocal and wind lines in the number 27a ('So ist mein Jesus gefangen'); these are not indicated in Altnickol’s manuscript but we have inserted them according to the stylistic conventions in the 1720s. The same is true of the length of bass notes in the recitatives. It was the practice in recitative to write long notes in the continuo part—it is less time-consuming for the composer (and uses less ink) and the players would be aware of the conventions of the day. If there was any doubt the composer/director could very quickly clarify his expectations in rehearsal. An interesting feature of the manuscript is the lack of full texts for the chorales—only the opening, identifying line is given. Also missing in the chorales is any indication of orchestration; we have taken this as an opportunity for performers’ decisions and performed some with voices and organ only, allowing a more vivid communication of the text.

The essential structure of the Passion is that with which we are familiar from later versions. The most noticeable difference is the chorale at the end of Part One which was subsequently replaced by the large-scale chorale-chorus ‘O Mensch, bewein dein Sünder gross'. There are other less immediately noticeable changes—a variant text for the chorale number 17 'Es dient zu meinen Freuden'—and matters of orchestration for some choruses such as the wind doublings in the openings chorus. More obvious and significant differences include the allocation of the solo line in the opening movement of Part Two to the first choir bass rather than to the alto and the absence of a gamba from the orchestral line-up. In this version a lute obbligato in the bass aria 'Komm, süsses Kreuz' gives a still greater intimacy to the moment; with the vocal performing forces Bach would have used (and employed here also) this has still greater poignancy in that it is sung by the same singer (Bass, Choir I) who sang the role of Christus. The basso continuo line in this aria is very similar to the later version but the gamba obbligato would have allowed a string accompaniment (possibly both cello and violone) alongside organ. The exquisite baroque lute would have been difficult to hear in a building the size of the Thomaskirche (hence its replacement in Bach’s later version); there are no such problems in recorded performance. We have taken further advantage of this medium by using harpsichord as the sole continuo instrument for this movement. The absence of a gamba from Bach’s scoring also affects the second tenor aria, number 35 ‘Geduld’, where the manuscript indicates merely a bass line, presumably a cello but perhaps also a violone either doubling or mapping out the harmonic structure. Otherwise, we might notice notation differences not only in the choruses but also in some solo vocal lines. This is most obvious in numbers 8 ('Blute nur') and 35 ('Geduld'); there are also minor differences in the word setting in such as number 13 ('Ich will dir') and others. There are further other surprising moments, at least to modern ears more familiar with Bach’s later settings, including some accidentals in the opening chorus, the bass-line chromaticism in number 8 ('Blute nur') and the tempo change for the second chorus in number 19 ('O Schmerz!').

The use of the harpsichord not only here but also in the secco recitatives might also surprise some, but commentators such as Laurence Dreyfus have made a convincing case for the availability, indeed the requirement by Bach, of a harpsichord in the organ lofts at the Thomaskirche as well as the Nicolaikirche. It seems likely that Bach used the harpsichord as a director’s communication aid rather than an essential harmonic element, but it offers an attractive element to the continuo group in recitative in which we again take advantage of Dreyfus’ information that Bach had not only the now more fashionable cello and organ but also violone and bassoon, forming a group of five instruments and allowing considerable variety of texture, dynamic and articulation to be used at the performers’ discretion.

It seems quite clear that Bach was able to employ what for him were relatively large-scale forces for his 1727 performance and although we occasionally now hear performances and recordings using these small numbers it is still nowadays more usual to hear double choirs and orchestras with independent solo teams. Bach had a team of four concertato (SATB) singers and a similar number of ripienists; these made up the two choirs in this Passion with the Tenor and Bass in Choir I singing not only choruses but also the relevant arias and the roles of Evangelist and Jesus. (For logistical reasons we didn’t quite manage to achieve this!) All the other chorus singers have arias also; the ripieno chorale 'O Lamm Gottes unschuldig' in the opening chorus was possibly sung by a small number of singers, probably as few as three (as here), or might perhaps even have been played only on the organ. The minor roles of the two maids and Pilate’s wife were sung by just one of these ripienists; of the two other male singers, one sang Judas and Priest I, the other Peter, Priest II and Pilate; the two false witnesses were sung by Choir II alto and tenor. Current scholarship suggests that the orchestral parts were similarly all performed one to a part for this early version, meaning that balance between woodwind and strings as well as between voices and orchestra was rarely an issue.

Whatever the differences between this early and his later versions Bach’s structure and dramatic purposes shine through. One of the most moving products of recent scholarship has been the realisation that a large chunk of this work, in common with Johannes-Passion and several other vocal works from this period, is constructed in a large-scale palindrome which, to those trained in baroque Rhetoric, would suggest a chiasmus or structural cross figure. The numbering of the score unhappily obscures the precise nature of Bach’s ground plan, but essentially the second part centres on an extraordinary celestial piece, scored for soprano, flute and two oboes da caccia without the usually ever-present basso continuo; that is the very obvious aural signal for Bach’s audience indicating something very special in the text. Working outwards from this point (No 49) we find a strict mirror arrangement of movements, representing a huge architectural cross built into the fabric of the work. This exists alongside other chiasmus figures: some visual/aural melodic figures in the shape of a cross such as the melodic (‘Z’ as in ‘Kreuz’) cross shape for the opening melodic material in 'Blute nur'—similar to Bach’s own name: B flat–A–C–B (B=H in terms of German nomenclature); the superimposed tonic/dominant harmonies at the start of the opening chorus; and some purely visual—the horizontal, contrapuntal lines of Choir I set against the vertical interruptions of Choir II in the opening chorus tell us the purpose of this monumental piece. The palindrome-structure chiasmus set around numbers 46 to 49 helps us to now identify the structural centre. The discovery of this structural chiasmus, an obviously private symbol, in the work is perhaps made all the more moving by its hidden nature. The original listeners couldn’t have been aware of it. Clearly it represents a private act of devotion; equally clearly it presents musical and dramatic difficulties to the composer for no apparent technical end other than as an act of particular personal dedication. The discovery of the chiasmus contributes immeasurably to our picture of Bach as a composer whose private sincerity often outweighed his public exterior.

It has been said, in a life of Martin Luther, that 'Bach was, by any sensible definition, the most influential evangelical of his age', and the composer’s intention here was a clear, direct reading of the Gospel story elevated to sublime expression. The texts themselves divide into two categories, the Gospel text, and the interpolated texts taken from Picander’s Erbauliche Gedancken of 1725, which again follow Luther’s dictum, aiming to elucidate the Gospel, and deliberately applying the historical story to the man in the pew. For example, the episode dealing with Peter’s denial splits into three basic units: the historical story retold by the Evangelist, an aria sung, as it were, by the personification of Peter, and a chorale giving the congregational response to the story; the whole procedure moves gradually from a statement of 'He did this' to a response of 'I do it daily, Father forgive me' (see nos 38–39). It seems unlikely that the chorales were in fact sung by the congregation but by the two choirs on behalf of the listener, but the sentiments expressed are always the response of the faithful Christian, experiencing Christ in his own situation.

Peter Seymour © 2015