Explanation for Variant Keys

No clear explanation for these variant keys is universally accepted. One theory that had been put forth suggests that the editors of the Birnstiel–Breitkopf publication are responsible for transposing the settings for practical reasons — e.g. to place the settings into a more comfortable range for average singers, to facilitate easier printing, etc. This theory is supported by a comment made by C.P.E. Bach in the prefatory note to the original publication of the Breitkopf edition in which he explicitly allowed for the transposition of settings to accommodate voice ranges. However, these considerations cannot fully explain all the key divergences in these collections. Four of the above settings appear twice in the Breitkopf–Riemenschneider collection, once in each key listed. The theory fails to explain why a key would at one time be deemed satisfactory and at another be deemed impractical. Furthermore, most of these chorales appear in these divergent keys in collections that predate the Breitkopf publication, in collections like the Fasch document from 1762 and source Am. B 46IIa from the 1770s, both of which are handwritten. Thus, the transpositions would have to have been exacted by copyists of earlier manuscripts and for reasons other than to alleviate any printing impracticalities.

Another possible explanation relates to the complex situation regarding pitch and tuning in Bach’s day. Instruments were tuned at two primary pitch levels — Chorton and Cammerton, the former generally being approximately one full step higher than the latter. In Bach’s Weimar, strings and voice were notated with the Chorton organ and trumpets 1 or 1.5 steps lower than the Cammerton woodwinds while in Leipzig, strings and voice were notated with the Cammerton woodwinds a whole step higher than the Chorton organ, trombones and (usually) trumpets. (See Beverly Jerold’s 2000 article, "Pitch in the Vocal Works of J.S. Bach" for a good summary of this very complex issue.) What this means is that in Leipzig era cantatas, the organ continuo parts were consistently notated a whole step lower than voice, string and woodwind parts. Could it be that the transposed chorales represented transposed Leipzig organ scores? This is certainly a possibility for the ten cases of settings scored a whole step lower, but what about the other settings at different transposition levels? It could be that the settings appearing a minor 3rd lower also represent transposed organ scores, as the minor 3rd transposition level was not uncommon in the Chorton–vs–Cammerton relationship. Still, this explanation is not entirely convincing. First, what are we to make of the fact that organ continuo parts generally consisted only of a figured bass line? Was there a separate "organ hymnal" from which these transposed Breitkopf settings were taken? Second, if the transposed settings were taken from such a hypothetical source, why are there so relatively few transposed settings?

Alternatively, it may be that the chorale settings from the these collections that appear in divergent keys actually predate the versions found in the large choral works in which they appear (i.e. cantatas, passions, etc.). NBA scholars have theorized that J.S. Bach himself had a collection of his own chorale settings, a hypothetical source they refer to as "[Y]". (See NBA III/2.2 KB, pp.84–85) The scenario envisioned here is that Bach consulted this source [Y] when composing his large choral works, transposing the settings himself to keys that fit the key scheme of each work. This theory is supported partly by the fact that the keys represented in these early collections are consistently simpler (i.e. fewer accidentals in the key signature) and in fact often match the keys in which these tunes are placed in the hymnals of Bach’s day, hymnals like the Neu Leipziger Gesangbuch, the suggestion being that the settings were originally composed in the keys in which they normally appear and are normally sung. Generally speaking, the transposed continuo parts discussed in the previous scenario more often than not fall into keys that are less common, while in the case of these transposed chorales, the settings in these early collections fall into keys that are more common — only four of the nineteen Breitkopf settings have key signatures featuring more accidentals relative to the key in which they appear in the large choral works, and only three settings in the Breitkopf contain more than two accidentals compared to the 12 such key signatures among the nineteen settings as they appear in the large choral works. It could certainly be argued from this evidence that the settings were transposed into simpler keys in preparation for these early collections, but such an argument loses force when considering that many other settings in the Breitkopf edition retained the larger key signatures Bach used. Why would some key signatures be deemed unsatisfactory in a few cases while being deemed perfectly satisfactory in many other cases? (On p.96 of NBA III/2.2 KB, scholars at least envision the scenario described here, if they are forced to conclude that it can "neither be proved nor disproved" (p.97).)

Of course, it is certainly possible that no one single explanation can explain all of these transpositions. At any rate, we are forced to conclude as the scholars of the Neue Bach Ausgabe have: "Whether editorial divergences from the original sources are attributable to interventions by a third party or whether they are derived from another source, such as J.S. Bach’s own collection of chorale settings, can neither be proved nor disproved." [Ob von den Originalquellen abweichende Lesarten auf Eingriffe von fremder Hand zurückzuführen sind oder ob sie aus einer anderen Quelle, etwa einer J.S. Bachschen Sammlung von Choralsätzen stammen, läßt sich weder belegen noch widerlegen. ] (NBA III/2.2 KB, p.97)

bach–chorales.com by Luke Dahn. Copyright 2017.