In addition to these nine libretti, both Spitta[445]and Schweitzer[446] attribute to her the text of Bach's Cantata for the Second Sunday after Easter in the same year.[447] It is uniform in construction with the authentic nine, but is not among the authoress's published works. Wustmann[448] [pg 181] finds the tone of the libretto less ardent and its rhythm rougher than those published under her name. Admitting the soundness of Wustmann's criticism, one hazards the opinion that the challenged text was written at the period when Bach set it, namely, in 1735, eight years after the poetess published her earlier texts. The difference of time may account for the difference of texture to which Wustmann draws attention, but leaves undecided the question whether Bach was drawn to the earlier through the later and unpublished texts or vice versa. It is quite probable that he set other libretti by the same writer, though Schweitzer's[449] attribution to her of a second text for Ascension Day, 1735, must be rejected.[450]
It is worth noticing, since it certainly reveals Bach's preference, that Marianne von Ziegler's libretti are constructed almost invariably in the Weiss form. Every one of them but three[451] opens with a Bible passage, invariably taken from St. John's Gospel, which provides the Gospel for the Day from the First Sunday after Easter down to Trinity Sunday, excepting Ascension Day. All but one (No. 68) of the libretti conclude with a Choral, and their Arias are hymn-like in metre. The tone of them, however, is warmer, more personal, less didactic than the Weiss texts. That Bach regarded them with particular favour is apparent in the circumstance that he took the trouble to revise all but one of them.[452] That they stirred his genius deeply is visible in the settings he gave them.
After 1735 the chronology of the Cantatas is not certainly ascertained. Of those that fall after the Ziegler year, as we may term it, the majority can only be dated approximately as circa 1740, that is, anywhere between 1735 and 1744. Nor, except rarely, can we detect in their libretti the work of those on whom Bach elsewhere [pg 182] relied. Weiss, who died late in 1737, is only an occasional contributor. The texts of this period, in fact, are the outcome of Bach's own experiments in libretto form. Thirty-three of them are Choral Cantatas, whose evolution it remains to trace concisely.
That Bach should have turned to Lutheran hymnody, chiefly of the fifteenth and sixteenth centuries, and that the Cantatas built upon it should be his most perfect religious work is not surprising. The hymns and their melodies were the foundations upon which the temple of German Protestantism had been reared. They appealed vividly and powerfully to Bach's spiritual nature, and profoundly influenced his musical utterance. His whole career, as Sir Hubert Parry points out,[453] was an effort to widen his means for self-expression. And the Choral Cantata, in effect, was the reconciliation or blending of this self-discipline. It was the supreme achievement of Bach's genius to assert the faith and idealism of Lutheran hymnody with the fullest resources of his technique.
It is not our task to consider the hymn libretto in its relation to the structure of Bach's latest Cantatas. Necessarily it tied him to a stereotyped design, which he clung to with greater persistency because it exactly fulfilled his devotional purpose. But experience compelled him, after a brief trial, to discard the simple hymn libretto. In the earlier Leipzig years as many as eight Choral Cantatas[454] are set to the unaltered text of a congregational hymn. In the later Leipzig period only two[455] libretti are of that character. Bach, in fact, soon realised that, while the unaltered hymn-stanza, with its uniform metre and balanced rhyme, was appropriate to the simple Choral or elaborate Fantasia, it was unmalleable for use as an Aria or Recitative. Hence, retaining the unaltered Hymn-stanza for the musical movements [pg 183] congruous to it, he was led to paraphrase, in free madrigal form, those stanzas which he selected for the Arias and Recitativi.
As early as September 16, 1725,[456] Bach was moving towards this solution. And it is significant that Picander's hand is visible in the libretto. The next example[457] occurs three years later, and again reveals Picander's authorship. Two other instances also occur in the early Leipzig period.[458] To that point, however, it is clear that Bach was not satisfied as to the most effective treatment of the hymn-libretto. But in the second Leipzig period, after his collaboration with Marianne von Ziegler, he arrived at and remained constant to a uniform design. Of the thirty-nine Choral Cantatas of the whole period only two exhibit the earlier form. Of all the others the libretto consists partly of unaltered hymn-stanzas—invariably used for the first and last movements, and occasionally elsewhere—but chiefly of paraphrased stanzas of the hymn, whose accustomed melody, wherever else it may be introduced, is associated invariably with the hymn when the text is used in its unaltered form. We, to whom both words and melody are too frequently unfamiliar, may view the perfections of the Choral Cantata with some detachment. But Bach's audience listened to hymns and tunes which were in the heart of every hearer and a common possession of them all. The appeal of his message was the more arresting because it spoke as directly to himself as to those he addressed.
It would be satisfactory and interesting to point positively to Bach's own handiwork in these libretti, of which he set fifty-four in the period 1724-44. Unfortunately it is impossible to do so, except, perhaps, in a single case,[459] [pg 184] where we can reasonably infer that the libretto is his. Of the rest, one is by Franck.[460] In eighteen of them the hand of Picander is more or less patent.[461] Nineteen[462] we can only venture to mark “anonymous,” though Picander is probably present in most of them. Ten are unaltered congregational hymns.[463] There remain, however, five[464] in which, perhaps, we detect another, and the last, of Bach's literary helpers.
Wustmann draws attention[465] to the libretto of Cantata No. 38, a paraphrase of Luther's Psalm 130. He finds in it, and reasonably, an expression of “Jesus religion” very alien to Picander's muse, and suggests the younger Christian Weiss as the author of it. Like his father, he was Bach's colleague, the godfather of his daughter, and undoubtedly on terms of close friendship with him. But if he wrote the libretto of Cantata No. 38, probably it is not the only one. The same note rings in four more of the Choral Cantatas,[466] which may be attributed tentatively to Weiss, though their ascription to Bach would be equally congruous.
Returning, however, to the seventy-two libretti of the later Leipzig period we reach this result: More than half of them (thirty-nine) are congregational hymns, all but two of which are of the paraphrased type in which we detect the work of Picander, Bach himself, and perhaps the younger Weiss. Of the remaining thirty-three original libretti Marianne von Ziegler heads the list with nine, and perhaps ten.[467] Bach follows with a problematical [pg 185] six,[468] Picander with five,[469] the elder Weiss with four,[470] Neumeister with one.[471] One text is taken from the Bible.[472] Another consists of a single stanza of a hymn by Martin Behm.[473] Five are by authors unknown or undetected.[474]
But, as was said at the outset, the attribution of particular libretti to individual writers is conjectural, except in comparatively few cases. Yet, unsatisfying as it is, this guess-work reveals with approximate correctness the extent to which Bach drew upon his own and other people's abilities for the texts he needed. Summarising our conclusions, we discover that about one-quarter (fifty-four) of the 202 libretti set by Bach between the years 1704 and 1744 were provided by the hymn-book. It is shown elsewhere[475] that all but eleven of them are taken from Paul Wagner's volumes. The elder Weiss comes next with thirty-three libretti. Bach follows with thirty, Salomo Franck with twenty-one, Picander with twenty (exclusive of his arrangements of Choral Cantata texts). Marianne von Ziegler contributes ten, Neumeister seven, Eilmar and Helbig two each, Gottsched and Martin Behm one each. Three libretti are taken from the Bible or Church liturgy. Eighteen remain anonymous.