Group K Manuscripts
The MSS sources originating from Kirnberger's circle survive in large number, in a wide diversity of corpus, format and textual authenticity. From this we learn that Kirnberger was not only interested in compiling an authoritative version of WTC II for Amalienbibliothek, but also in wider circulation of the work.
The ultimate model, K', does not survive. This study suggests that it was compiled from various sources: it is most likely that the majority of the movements are based on H(possibly the personal copy that Kirnberger made under Bach in 1738-41), and the p.corr. reading of L (most likely a set of copies he later obtained for this purpose, which is related to P 580, P 582, etc.) was later overlaid on the top, at an early stage of Kirnberger's own compilation of WTC II. The exceptions to this are PrFg.f, which is based on the earliest known version of the piece which Kirnberger must have acquired from Bach while studying in Leipzig, and PrFg.C and Ab, which stem from L p.corr. This method of working seems to have been maintained throughout the development of its singular tradition, for some of the unique errors remain in the descendant MSS of K as a hallmark to this (such as Fg.f#, b.44). The genealogical tree of K, with noteworthy eclectic revision activities between sources, is reconstructed in the following diagram:
From the earliest stage of K' stems the fugue only source, RCM 26, and subsequently, its direct copy Cambridge. Kirnberger then acquired a correction exemplar belonging to A1. There seems to have been at least two separate occasions when Kirnberger used a MS from A1 as correction exemplar. The first occasion constitutes the major operation, including the extension to Fg.e. Thus all the remaining groups of sources stem from this contaminated state of K'. The next distinguishable revision using A1 seems to have taken place before Am.B.57, the so-called "Kirnberger's Handexemplar", was made (e.g., Fg.C, b.13). This minor revision only affects the MSS of K1.
It seems significant that before Am.B.57 was made, there survived only 'incomplete' copies of WTC II. Although the actual chronology of these sources is yet to be proven, the text-critical data seems to point to circumstances where Kirnberger was not ready to make a complete volume, possibly still lacking some movements such as Pr.Ab. From this observation, the character of K2 and P 211 can be described as a pre-K1 stage of compilation, and that K3', a lost common source for the non-authoritative branch of K, stemmed from stage K1.
Once Am.B.57 was made, Kirnberger decided to enter his own inspired readings into it, e.g., Fg.f, b.64. This type of revision appears to have continued from time to time even after copies were made from it. On other separate occasions, Am.B.57 was corrected together with P 211.
Sometime later, Weyse was copied from Am.B.57, possibly directly, by the same scribe, Anon.402. Its main purpose seemed to be to rewrite the volume in more spacious format. A large number of pitch errors were made, probably due to the mechanical nature of the task. There is evidence of collective correction being carried out between Weyse and Am.B.57 (and often P 211), followed by the systematic correction of Weyse alone. Many ties and rests were added (i.e., restored) and the shape of appoggiaturas was refined. Unfortunately, many of the pitch errors remained uncorrected. This state of Weyse is characterised by K1ii. From this state of Weyse stemmed Am.B.49, this time copied by Anon.403. Weimar was also made from K1ii, but is full of errors. Similar collective corrections were made at this stage also.
At some stage after all of the different Groups of sources were clearly distinguished and established, further collective minor revisions were made, e.g., Pr.d: b.11; Pr.f: b.32: A,1/2.
Finally, at yet another time in Kirnberger's continuing revision processes, a very important amendment to his policy was made, so that many valid, often more correct, readings which deviated from L began to disappear. This extraordinary decision is best justified by the hypothesis that Kirnberger encountered Bach's autograph, L.9 While Am.B.57 was hardly changed (except for Fg.C#, b.29: S,4 and Pr.d, final cadence), K1ii went through revisions collectively (e.g., Pr.c, b.14: S and Pr.E, b.50: B,1). It is difficult to place this event chronologically in the series of revision activities, for it appears in various patterns, some as early as the pre-Am.B.49 stage.