Bach: Music in the Castle of Heaven

by Gardiner, John Eliot

  We have now arrived at the apex of Bach’s nine-movement Gloria. At this point he finds ways of bringing to the surface that partially concealed narrative thread of the Mass Ordinary (which eludes all but a few composers). You sense his delight in juxtaposing a Gregorian-based movement like the majestic Gratias with an elegant galant duet like the Domine Deus, which uses the same rising bass line (three whole tones and a half step) in double diminution and follows it with a similar melismatic figure, proof that with his technical mastery he can encompass any mood or style at will. Next he pairs the Domine Deus duet (without its expected da capo structure) with the Qui tollis chorus. The dramatist in him makes delicious play of the seraphic innocence of the filial relations between Father (tenor) and Son (soprano) in a canonic duet. This he contrasts with the extreme pathos of the Qui tollis – the pain caused to the Son-made-man in anticipation of Christ’s suffering on the Cross. Everything about the Domine Deus is conceived in terms of benediction: the key (G major), the benign mood created by flute over pizzicato bass and answered by muted upper strings, the sense of its being a spiritual love-duet – both modern and galant. Even the two eldest of Bach’s fashion-conscious sons might have approved. But their father is not playing to the gallery. If the task is to find suitable music for such a central doctrinal text, he seems to be saying, then set it with a smile – with euphonious parallel thirds and sixths, syncopated or plain. Musicologists have been oddly perplexed by the clear traces of rhythmic alteration, the back dotting of the paired semiquavers to be found in the Dresden part-books. That Lombard rhythms (back dotting) were fashionable in Dresden in the 1730s, particularly in movements which emphasise Christ’s intimate relationship to mankind, does not imply that Bach was currying favour with the Dresdeners; it is merely the gentlest of hints that he was au fait with current trends, one likely to have found favour with the likes of Zelenka and Pisendel.k More to the point, they fit idiomatically with the sentiment of the words and give lilt and charm to the flute theme, with its answer in the strings.

  Now, with the words Domine Deus, Agnus Dei, a shadow passes across the music as it modulates to E minor, as though in anticipation of the Crucifixion. It is also perhaps an indication to the performer that the back dotting should now cease. If this was Bach’s way of indicating to us a B, or a ‘middle section’, he confounds our expectations by doing away with a da capo altogether and merging his duet imperceptibly into the ensuing four-part Qui tollis. He then slows the momentum and eases us into the sarabande-like chorus, even adapting the outline of the final phrase of the singers in anticipation of the melodic shape of the new sorrow-laden chorus proclaiming that God in Christ bore the sins of the world on the Cross. This is the central text of the Gloria and, by means of this reminder, its most serious moment. As the ‘miserable offenders’ necessitating Christ’s atonement, we are referred back to those desperate cries of help at the outset of the Kyrie. l

  The prevailing mood of this remarkable fifty-bar movement is one of woebegone heaviness and anguish, generated by the violas, now un-muted. Their paired quavers slice through the texture with sighs of lamentation (just as they did in the opening chorus of the John Passion). They are the grieving heart at the centre of the choral and orchestral body. The clarity of the vocal scoring – initially two voices at a time – makes this grief appear to be more personal. A pair of flutes enters high above the dark sonority of voices and strings, a soothing and occasionally disturbing presence. Bach has chosen his Jeremiad cantata (BWV 46), the one predicting the destruction of Jerusalem, as his model (see this page), crucially adjusting its rhythmic inflection – and he makes these adjustments straight on to the score. The measured declamation of wie mein Schmerz (two crotchets and a minim stretched over a descending diminished fourth) now becomes the much more urgent Miserere nobis (in four consecutive quavers) – a tiny, innocent-looking adjustment, but huge in its new expressive potency. Significantly he also omits the cantata’s opening sixteen bars of instrumental prelude, plunging straight into the opening Qui tollis.m

  There are several ingredients to this remarkable creation. First, there is the background of the familiar Old Testament penitential text, with its reference to the destruction of Jerusalem as predicted by Jesus (Luke 19:-41–8). This is conveyed through the harmonic dissonance and intense expressivity of the vocal lines. Then there is the single emphasis by the bass line at the start of every bar; and the hovering of those flutes – serene at first, but later fluttering like wounded birds. The effect is poignant (still more so in our own times, when the implied references to Jerusalem and the frequent threats to its sites, holy to three religions, are excruciatingly topical). The individual vocal lines begin in imitation. Sometimes they collide, with prolonged bruising dissonances, then part, each apparently pursuing an independent trajectory. There are momentary pairings (deprecationem nostram), and all four voices come together at cadences. This gives to the whole movement a sense of tragic choreography – of slow-motion spirals, of furrows being turned, or of geometric dance patterns loosened and redrawn. Only the highest and lowest voices bind things together: a single harmonic emphasis per bar in the continuo (but pulsated in the cello line like a slow bow vibrato) and the flutes high above the enactment of this pained human ritual. We may be tricked into hearing this amazing polyphony as the result of autonomous melodic movement, whereas all the time it is being controlled by the inexorable harmonic rhythm, the tonal grammar of Bach’s bass line. In performance the effect of such a tableau vivant etched in sound is spellbinding – provided that every one of its ten strands (four voices, four strings and two flutes) is balanced, combed and always distinct.

  Bach has ended on a half-close. In a Passion-setting one might expect a secco recitative at this point. In Bach’s original (BWV 46) the music now erupts into an energetic fugue. Here, on the other hand, he keeps things moving forward by means of consecutive arias. Once again, as with the Kyrie trilogy, we encounter the rubric Qui sedes sequitur and Quoniam tu solus sanctus sequitur. The music is through-composed, in other words, and no awkward pauses are called for. Doctrinally Bach has moved from Atonement to Mediation. In the first of the two arias, the Qui sedes (for alto, oboe d’amore and strings), he portrays Christ’s role mediating between God and man, and ‘sitting at the right hand of God’. This is symbolic and even ironic; for here the music is anything but sedentary – as an Italianate giga it is unequivocally balletic and its ritornello structure is made up of overlapping phrases which erode the stability of its underlying dance pulse.

  In the second aria (Quoniam) the text refers to Christ’s kingly office. It clearly appealed to Bach’s particular vein of humour to evoke the ‘most high’ with the growliest forces available to him: two bassoons, bass soloist and basso continuo. The treacherous Waldhorn is the exception here. The bassoon counter-theme, as Tovey rightly insists, ‘must always be brought out as a main theme and not treated as an accompaniment’,7 particularly at the point where they chug along above the horn line in buffo style (bars 72–4). It seems that he had the sound of a particular ensemble, even of particular individuals, in mind. Five bassoonists were on the payroll of the Dresden Capelle, and two or three of them were on call at any given time (whereas in Leipzig he was lucky to find even a single competent bassoonist). Both Heinichen and Zelenka regularly composed for pairs of solo bassoons, and Hasse had featured the Waldhorn in his opera Cleofide, which Bach heard in Dresden in 1731. In Bach’s hands the overall effect of these instruments in combination is magnificently stilted, bucolic and slightly grotesque. As befits a polonaise, the horn is noble, regal even (but we need to ensure that the singer is never engulfed by the surrounding sonorities), while the continuo line often bumps into – and sometimes rises above – the bassoons.n

  The Quoniam begins to make sense by its strategic placement, in the way that it follows the pathos of the Qui tollis and its dance-like sequel, and heralds the epic razzmatazz of the Cum sancto spiritu – the three movements are closely bound tog
ether. Bach reminds us that this is the completion of a clause – ‘for thou only art holy … with the Holy Ghost in the glory of God the Father. Amen.’ Towards the end of the aria there is palpable expectancy as the bass-messenger rounds out his proclamation and the instruments pipe him off stage. Immediately the Cum sancto takes off with a tremendous jolt, rather like the way a Big Dipper deceptively inches its way along, then suddenly hurtles off. Mention of the Holy Spirit is the key to the changes in both pace and mood of this new music. As in his double-choir motet BWV 226, Der Geist hilft unser Schwachheit auf, the invigorating power of the Holy Spirit is the determining factor in the Christian’s acknowledgement of Christ’s Godhead. This is cue for celebration – in dance as much as in song. The throbbing pulsation of a single note (which was already present in the horn part of the Quoniam) passes to the upper strings, and later the trumpets and finally the woodwind, but now with a greater sense of forward propulsion and ebullience. For this is the fourth movement in a row to use triple or compound metre, and in ever-quickening tempo: the Qui tollis as a sarabande, the Qui sedes as a moderately paced giga, the Quoniam as a stately but forward-thrusting polonaise, and finally the Cum sancto, a free-spirited, corybantic dance.o

  The sense of release and of liberation in this fabulous choral dance is contagious. Bach’s technique is at first to alternate contrasting groupings of voices and instruments to build structure and generate excitement. The fleet-footed exposition of the fugue then begins in the voices alone (bar 37) and bears a striking affinity with the balletic choral fugue ‘Die Kinder Zion’ from his motet BWV 225, Singet dem Herrn (see this page). Now the instruments re-assert themselves – downward arpeggio cascades in the strings, a jazzy syncopated figure in the winds, and skittish curlicue flourishes which are there purely for high spirits (for they do not advance the thematic argument one jot). The orchestra wills the choir to re-join them with their plain, chordal Amens. On their next appearance the vocal lines are doubled by the instruments in a second fugal exposition (which, interestingly, shows slight variants, hinting that a four-voiced original may stand in the background of this movement’s creation). The trumpets are finally sucked back into the action, and the music breaks free with the sort of Dionysian abandon one might associate more with Beethoven or Stravinsky than with Bach.

  We have seen that the Lutheran year had its feasts no less than its fasts, and again and again in the cantatas Bach delights in the seasonal punctuation of the year and in any of the pagan festivals that Christianity appropriated for its own calendar. Well, here is one – let us say it is for Midsummer’s Day – which was not included, or as far as we know given official approval. It is celebrated in carnival style – ‘almost pre-Christian, if not overtly pagan, in its abandon’.8p The contrapuntal zest that Bach generates in these final bars – and the aural pleasure he gives us – is immense. Part of its magic lies in the several ways he finds of dividing the twelve semiquavers of a bar into different groups, involving cross-rhythmic patterns and syncopations.

  What, then, did those Dresden Court musicians make of this ‘apotheosis of the dance’? We have seen that there was much that would have been familiar to them in the earlier movements of the Gloria – the sectional treatment, the balancing of solos with choral movements, the florid writing for solo voices and obbligato instruments – which displayed Bach’s perfect assimilation of modern Neapolitan style exported to the Saxon capital. But a choral movement with this degree of athleticism and secularity was surely something utterly new. The contrapuntal virtuosity was dazzling enough, and the uncompromising exuberance of his writing for voices and instruments – particularly for trumpets – was beyond anything in their repertoire. Perhaps for at least one of them, Zelenka, the spicy rhythmic virility and ornamental daring of the Cum sancto put him in mind of his native Bohemia.q

  For the next twelve years we lose all trace of the Missa and its potential expansion. Was it dead in the water, as it were, after the big disappointment in Dresden? Did Bach just file it away in the well-stocked recesses of his memory bank, waiting for a new set of circumstances to present him with an opportunity for revival and reappraisal? If so, the single trigger that detonated the creative energy needed to complete the Mass is probably to be found around Christmas-time in 1745. The Second Silesian War had just come to an end, having brought considerable hardship to Leipzig and its citizens. For the first time in his life Bach had first-hand experience of the horrors and suffering of war, as Prussian troops occupied Leipzig in the late autumn of 1745 and devastated its surroundings. Three years later he still remembered it as ‘the time we had – alas! – the Prussian invasion’.9 A special service of thanksgiving to celebrate the Peace of Dresden was held in the Pauliner (University) Church on Christmas Day. Sandwiched between the early morning Mass in the Thomaskirche and the afternoon service in the Nikolaikirche, this was one of those occasions when members of Bach’s two best church choirs were available to perform together.r Here also was an opportunity to give Leipzig audiences a chance of hearing his unusual five-voiced Latin cantata BWV 191, Gloria in excelsis Deo – in which he had hastily re-assembled and condensed three of the Dresden Gloria movements (Gloria, Domine Deus, Cum sancto spiritu) into a new triptych. In addition, his six-voiced Christmas Sanctus, first heard on Christmas Day in 1724, was almost certainly revived for the same service. So, five of the eventual twenty-seven movements of the B minor Mass might have been performed together for the first time. Given the political context and the sense of collective relief at the ending of the war, it is possible that we have here the embryo of a Friedenmesse, a ‘Mass of peace’.10 That would have been consistent with the in-built alternation of human woe (Kyrie) and its release in God-inspired joy (Gloria) latent in the structure of Bach’s Dresden Missa. Was he struck afresh by the quality of his Latin-texted music? Perhaps he suddenly saw a destiny for it – she potential for incorporating it into a much more ambitious framework, one which gave him the motivation to create a definitive statement of faith comparable in scale and grandeur to his Passion-settings.

  At some stage, then – perhaps immediately after the Christmas peace celebrations, but perhaps not for another two years – came the momentous decision, with the Missa of 1733 as its starting-point, to complete his ‘Great Catholic Mass’. (This was the title under which it appeared in the estate catalogue of C. P. E. Bach in 1790.) The ramifications were huge. Bach was still in possession of the dedicatory score – but not the parts of the Missa. He would have reasoned that any complete Mass-setting would need to match the original Kyrie and Gloria pairing in terms of scale and structure. It meant that for the Credo, for example, he would need to devise a multisectioned movement of comparable weight to the preceding Gloria. The risk here was that he would end up with a work on a gargantuan scale, too long, apart from the most exceptional circumstances, to fit into any liturgy, no matter whether it was Catholic or Lutheran. Whereas his Kyrie/Gloria pairing was just about within the acceptable liturgical dimensions of the Dresden Mass repertoire, its new portions would be far in excess of what was considered fitting as a setting of the Ordinary of the Mass.s Grouping them in this way for practical use in no way compromised what may have been his twin aspirations in completing the work: to encompass within a single work an encyclopedic survey of all the styles he most cherished in the music of his own and of earlier times, and to achieve perfection in the execution of that work. It shows phenomenal ambition.

  His preparations were meticulous, characteristic of the exercises he deliberately undertook every time he committed himself to formulating a definitive statement. Bach was reverting to first principles (in the same way that all good scientists do at some point), wanting to move beyond the limits of what he had done before. If that meant going back almost to square one, that would be a sacrifice worth making: he had decided to carry out a fundamental reappraisal of the very building blocks, mathematical, musical or otherwise, that would enable him to refresh and to bring alive his thoughts and quest
ions about the Mass and its wider implications.

  First came considerations of basic structure, logistics and style. The choice of the Missa of 1733 as the starting-point for a new Missa tota meant that some aspects had already been settled: the five-part vocal scoring and the full-sized orchestra; the division of the text by coherent (though linked) sections fitted into a chiastic structure; the mosaic of solo and choral movements and the intermixture of styles – Italianate concertato movements on the one hand, and pronounced contrasts of archaic polyphony on the other. It is conceivable that Bach could have departed from the Dresden-specific style of the Missa at this point – but to what end? It had served him well so far and, besides, he had promised to furnish the Elector with further examples of his ‘indefatigable diligence in the composition of Musique for Church as well as for Orchestre’.11 As things stood, he might have felt that his best chances of obtaining a performance of so ambitious a setting of the complete Mass Ordinary lay in Dresden, not in Leipzig.