This article was originally part of a Round Table discussion entitled Four Approaches to the Understanding of a Single Musical WorkThe Aria "Ich folge dir gleichfalls" from the St. John Passion of J.S. Bach, which took place at the seventh annual meeting of the Society held in Washington, D.C., December 28-30, 1964.

The other participants were Donald J. Grout, Edward T. Cone, and Edward A. Lippman. Their articles also appear in SYMPOSIUM Volume 5.

 

The earliest surviving sources for the St. John Passion consist of

	A.	A score, of which the first 20 pages are autograph (No. 1 through No. 14, m. 42) and the remainder in the hand of a copyist whom Dürr1 calls Hauptkopist H. BG is based on this score.
	B.	An almost complete set of parts, including several duplicate parts. BG used these parts only as supplementary sources, where the score was ambiguous or imperfect.

A.	Of the score, pages 1-8 have one watermark; pages 9-24 have a second watermark; pages 25-92 have a third, or at least seem to have; at any rate, pages 25-92 are all on the same kind of paper, which is not the same as that of pages 1-8 or 9-24.
B.	The parts are on five different kinds of paper (none of them the same as any of the paper in the score). We may group the parts by watermark approximately as follows (Groups I, II, and III apparently each correspond to one performance; Groups IV and V seem to have been both prepared for a fourth performance):

	Group I (1724):	Violin I, Violin II, Soprano ripieno, Alto ripieno, Tenore ripieno, Basso ripieno, Continuo (untransposed and unfigured). All these parts are in different handwritings. The musical contents are, except in detail, the same as those in the score.
	Group II (1725):	An almost complete set of vocal and instrumental parts, lacking only a figured and transposed continuo part—nearly all of them in the handwriting of Johann Andreas Kuhnau, a student at the Thomas-Schule in the years 1718-1728. The musical contents of this set differ from the score in respect to Nos. 1, 19, 31, 32, and 68, and they contain an extra movement between Nos. 15 and 16.
	Group III (1728-1732):	Inserts, mostly autograph, originally sewn into the parts of Groups I and II. In 1725, the parts of Group I were modified to agree with those of Group II. Movements 1, 19, 31, 32, and 68 were crossed out or bracketed, and references were inserted to the substitute movements, some of which must have been present on inserts now missing. The inserts of Group III restored most of the original contents to Group I, and supplied them to Group II where their places had been occupied by other pieces.
	Group IV (after about 1747):	Violin I (a third copy), Viola (a second copy), and two Continuo parts, one figured, both untransposed. Violin I of this set is in the hand of Hauptkopist H, who wrote pp. 21-92 of the score. The other parts are in various hands, none of them the same as in Groups I, II, or III.
	Group V (after about 1747):	Inserts, in various hands, including Bach's, some of them the same as those represented in Group IV. At the time of the third performance, Nos. 61 to 63 had been deleted; and No. 19a, the replacement of No. 19 in the 1725 performance, had been in turn replaced by an aria which has disappeared. The inserts of Group V restored Nos. 19 and 61-63. They also restored No. 68, which had been replaced in 1725, and not restored by inserts of Group III.

 

*

*

*

Bach's handwriting in the autograph portions of Group III is very similar to that in the autograph portions of the score. In the autograph portions of Group V, it is strikingly different—it has a stiffness which we may perhaps attribute to rheumatic or arthritic troubles, or which could have been caused by long disuse of the writing hand, owing perhaps to some accident or illness.

This makes it possible for us to date with some confidence autograph additions and corrections in the parts of Groups I and II, which is important for our Aria.

This Aria occurs in the following sources:

A, the score.
Group I: B 21 (Unfigured Continuo).
Group II: B 1, 2 (Flutes), 14 (Soprano), 22 (Unfigured Continuo).
Group IV: B 23 (Figured Cembalo), 24 (Unfigured and hopelessly inaccurate copy of B 23).

The Aria is in Bach's handwriting in the score; it is in the hands of various copyists in the parts. In the score it reads almost exactly as in BG. In all the parts, the original reading is almost exactly as in the manuscript copy that has been distributed.

The parts, B 1, 2, 14, 21, and 22 all are earlier than the score; the readings of the score are improvements over those of the parts. B 23 and 24 are later than the score, but, like the rest of the parts, they contain the earlier readings. How are we to account for this?

Remember that when the 1725 parts were copied some of the pieces in the 1724 parts were crossed out, and when the inserts of Group III were written, around 1730, these pieces were restored in the 1724 parts and supplied in the 1725 parts. All these parts go back to an original score which does not survive. And that score must have been, by 1732 or thereabouts, heavily marked up, for in it some pieces had been first crossed out, in 1725, and then restored in some fashion, about 1730.

In the later 1730's, then, Bach must have planned a fourth performance and begun to make a clean copy of the score. As usual when Bach copied a piece, he could not resist the desire to revise it. But apparently, for some reason he dropped the project, and never resumed his copying. On March 17, 1739, the Clerk of the Town Council recorded that he had informed Bach that "the music he intends to perform on the coming Good Friday is to be omitted until regular permission for the same is received. Whereupon [Bach] answered: it had always been done so; he did not care, for he got nothing out of it anyway, and it was only a burden; he would notify the Superintendent that it had been forbidden him; if an objection were made on account of the text, [he remarked that] it had already been performed a couple of times. . . ."² Arnold Schering³ observed in 1941 that the Passion referred to was probably the St. John. This guess must have been based on the fact that according to the rules of the Town Council, the Passion was sung in one year in the Thomas-Kirche and in the other year in the Nikolai-Kirche. We know that in 1739 it would have been the Nikolai-Kirche's turn.⁴ The choir-loft in that church was too small to accommodate the St. Matthew Passion.⁵ This does not necessarily mean that the Passion Bach intended to perform in 1739 was the St. John, nor do we know definitely that the Council did not end up by permitting the performance. But it may well be that the interruption in Bach's copying of the score and this incident are connected. The only dated occurrences of either of the watermarks contained in the autograph portion of the score are in 1738-1739.

At any rate, when Groups IV and V were copied, Groups I, II, and III—a large body of material—were already in existence, containing the earlier readings. Improved readings existed in the autograph portion of the score, but it was apparently decided not to take the trouble to recopy the first fourteen numbers in Groups I, II, and III, but rather to copy the same old readings into Groups IV and V, so that they could be used with the old parts. So for this performance the score, containing improved readings in its first twenty pages, must have been used along with the parts, which contained the old readings.

Only in a few places did Bach bother to try to reconcile the parts with the score, and this Aria is one of those places. In it, he not only entered most of the improved readings of the score into the parts, but he went further and changed the text in the soprano part. The changes he made are indicated in the manuscript reproduction that has been distributed, below the original text, which is the one printed in the Bach-Gesellschaft. (Why he made these text changes is a question we had better reserve for the discussion period. They were not made in the score.) Both the corrections in the musical readings and the changes in the text are in Bach's stiff, labored handwriting that is characteristic of Group V.

Thus the history of this Aria, so far as we can determine it from the extant sources, is:

	1724:	A performance for which only the unfigured continuo (B 21) survives, containing, of course, the earlier readings.
	1725:	A performance for which the flutes, soprano, and a second unfigured continuo survive (B 1, 2, 14, and 22), all containing the earlier readings.
	1728-32:	A performance in which the Aria was performed from the materials of 1724 and 1725 (the inserts of Group III do not affect this Aria).
	1739?	(At any rate between 1735 and 1742 [Dürr] or 1736 and 1740 [Dadelsen]6): An intended performance for which the only evidence is the autograph portion of the score, containing the improved readings.
	After about 1747:	A performance with the improved readings, and with the changed text. The only parts dating from this time are the figured cembalo part and its useless copy (B 23 and 24). Both originally contained the earlier readings, and in both the improved readings were inserted as corrections.

---

¹Alfred Dürr, "Zur Chronologie der Leipziger Vokalwerke J.S. Bachs," in Bach-Jahrbuch, 1957.

²Hans T. David and Arthur Mendel, Editors, The Bach Reader, New York, 1945, pp. 162-163.

³Johann Sebastian Bach und das Musikleben Leipzigs im 18. Jahrhundert; der Musikgeschichte Leipzigs dritter Band, von 1723 bis 1800, Leipzig, 1941, pp. 174f.

⁴Cf. Bach Jahrbuch, 1911, p. 51.

⁵Cf. Arnold Schering, Johann Sebastian Bachs Leipziger Kirchenmusik, Leipzig, 1936, pp. 144-152.

⁶Georg von Dadelsen, Beiträge zur Chronologie der Werke Johann Sebastian Bachs, Tübinger Bach-Studien, Heft 4/5, Trossingen, 1958, p. 112.