So how many Bach four–part chorales are there?

What seems to be a straightforward question is, in reality, rather complicated. One thing is for certain: the answer is not 371 or 389.

First, it’s important to clarify that we’re trying to determine the number of extant chorale settings. Since speculations regarding the number of lost church cantatas vary considerably, it is entirely impossible to know how many Bach chorales have been lost.

Okay. So how many extant chorales are there? The answer most likely falls somewhere between 390 and 420 depending on how they are counted. And here the complications begin. There are 419 Bach chorale settings that have been assigned separate BWV numbers, and there is one additional setting attributed to Bach (see point 9 below) that has not yet been assigned a BWV number. The complications occur when considering which of the following settings, all of which are included in the 420, should be eliminated from that total number.

1) Nine duplicates appear within the Schmieder catalogue, six of which constitute instances of Bach reusing a setting for a separate choral work (BWVs 41.6=171.6, 75.7=100.6, 59.3=175.7, 244.15=244.17, 12.7=69a.6, 120a.8=137.5) and three of which resulted from the erroneous assignment of BWV numbers to settings from elsewhere in the Schmieder catalogue (BWVs 253=414; 197a.7=398; 245.5 1st version=416). Of the six in the former category, one is identical in every way (75.7=100.6), four are identical but transposed to different keys (12.7=69a.6, 41.6=171.6, 120a.6=137.5, and 244.15=244.17), and three are identical but with slight variation in the obbligato instrumentation (12.7=69a.6, 59.3=175.7, and 120a.6=137.5).

2) Two slightly different versions of two chorales from the St. John Passion exist (BWVs 245.3 and 245.5), the first versions of each coming from the premiere performance of the SJP and the later versions coming from a later repeat performance.

3) Two settings from the BWV 253–438 individual chorales are slightly reworked versions of chorales from the cantatas (BWV 359=154.3, 279=158.4).

4) Two settings (BWVs 27.6 and 43.11) that are taken directly from the 1682 Neu Leipziger Gesangbuch (NLGB), an important hymnal of Bach’s day.

5) One setting (BWV 8.6) is a revision of a chorale setting by Daniel Vetter.

6) One setting (BWV 262) is a four–part adaptation of a three–part setting by Pachelbel.

7) Two settings (BWV 145a and BWV 272) are attributed to Carl Philipp Emanuel Bach in the newly rediscovered Fasch manuscript (SA 818). The former setting has also been attributed to C.P.E. Bach in two other early manuscripts, D–Bsa SA 817, Fas. 2 and B–Bc 16083 MSM.

8) Two additional settings (BWV 500a, 1084) come from the St. Mark Passion Pasticcio, the authenticity of which settings is questioned (and which I personally have come to doubt due to divergences from the Bach chorale style), despite the fact that they have been assigned BWV numbers.

9) A third setting from the St. Mark Passion Pasticcio ("O Traurigkeit, o Herzeleid") may very well have been composed by Bach, though it has not yet been assigned a BWV number.

10) Finally, nine settings from Bach’s output may not qualify as chorales in the truest sense of the word. This is particularly true of the four–part second movement from the motet Komm, Jesu, Komm! (BWV 229.2), which is certainly chorale–like (apart from the rather disjunct melodic contour), but which features a melody that is almost certainly composed by Bach himself and, thus, is not taken from the Lutheran chorale tradition. The eight other settings that feature melodies attributed, to varying levels of confidence, to Bach himself, are BWVs 248.42, 299, 315, 345, 357, 384, 400, and 1122.

Below is a table summarizing these settings.

So which of these 30 settings should be eliminated from the 420? The reader is left to answer, and that answer may depend on the reasons for counting them. For the scholar studying the evolution of Bach’s chorale style, all revisions of chorale settings are of primary interest, perhaps warranting the counting of such variants separately. On the other hand, if we were to tweak the question just a bit and ask how many entirely unique chorale settings were most certainly composed (mostly) by Bach, thereby eliminating 1) instances of Bach himself reusing a setting, 2) instances of duplicate appearances in the Schmieder catalogue, 3) other instances of variant settings, 4) the two instances of Bach using settings from the NLGB, 5) the four settings that are (at least to a significant degree) attributed to other composers, and 6) the two St. Mark Pasticcio settings of dubious authenticity, then 20 settings can be eliminated, lowering the total from 420 to 400.

The Question of Authenticity

However, if we were to pursue the question of authenticity to its fullest extent, other factors are worth considering. For example, for how many Bach chorales do we have original or primary sources? In other words, for how many of these 420 chorales do we have in either Bach’s own hand or in the hand of one of his Leipzig copyists? The answer to this question is somewhere in the neighborhood of 228, a number that only slightly reaches half of the overall total! This figure includes six of the nine duplicates mentioned in point 1) above, as well as points 2), 4), 5), 8) and 9). Eliminate these and the number of entirely unique settings that came directly from Bach’s pen (or from the pen of one of his personal copyists) is only 214! This means that 192 chorales in this collection have survived by way of secondary sources and, therefore, should be considered slightly less authentic.

But even all of these 192 do not attain a consistent level of authenticity. It is important to know how these 192 chorales have survived to understand why this is. These individual chorales have survived by way of two important secondary sources. The first is a collection of 149 Bach chorales handwritten by Ludwig Dietel around the year 1735 in Leipzig when Dietel was a student in Bach’s Thomasschule. The fact that scholars are almost certain that Dietel copied these directly from original manuscripts of large choral works (e.g. cantatas, passions, etc.) gives this collection a certain amount of primacy among the secondary sources. However, two unfortunate facts regarding the Dietel collection remain. First, Dietel included no original texts or background information for each chorale. Second, Dietel was not a terribly accurate copyist and numerous errors pervade his document. These two deficiencies are especially unfortunate in regard to the 49 Dietel chorales that originally came either from larger choral works that are no longer extant or from other hypothetical didactic sources.

The second important secondary source is a collection of 371 Bach chorales posthumously published by Breitkopf in the 1780s and edited primarily by Bach’s son, C. P. E. Bach, the source on which the ubiquitous Riemenschneider edition is based. The primary problem with this document in regards to the question of authenticity lies in the fact that C.P.E. Bach took considerable editorial liberties, perhaps for the purpose of enhancing his late father’s legacy. To take but one example, several instances of consecutives fifths in Bach’s chorales were "corrected" by Emmanuel or other early editors. These "improvements" are clearly shown in a comparison of Breitkopf settings against their appearances in Bach’s original manuscripts. A secondary problem with this Breitkopf document is the lack of text and contextual information (much like the Dietel).

With these deficiencies in mind, one could well argue that those of the 192 individual chorales that appear in the Dietel collection attain a slightly higher level of authenticity A) given that the level of certainty that Dietel copied these directly from original manuscripts, and B) given that C. P. E. Bach took such editorial liberties with the settings. However, prioritizing the Dietel over the Breitkopf settings requires careful engagement with the errors that Dietel made. This would also mean that the approximately 141 settings that have survived only from the Breitkopf (i.e. settings that do not appear in the Dietel) should be approached with a slightly lowered sense of authenticity.

So there you have it. The answer to the question posed in the title of this brief essay is somewhere between 216 and 420, depending on many factors.

bach–chorales.com by Luke Dahn. Copyright 2018.