Kellner's Copy of Bach's Sonatas and Partitas for Violin Solo
Russell Stinson Early Music, Vol. 13, No. 2, J. S. Bach Tercentenary Issue. (May, 1985), pp. 199-211. Stable URL: http://links.jstor.org/sici?sici=0306-1078%28198505%2913%3A2%3C199%3AJKCOBS%3E2.0.CO%3B2-8 Early Music is currently published by Oxford University Press. Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/oup.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact support@jstor.org. http://www.jstor.org Sun Nov 11 11:59:01 2007 RussellStinson J.P.Kellner'scopyofBach'ssonatasandpartitas forviolinsolo Johann Peter Kellner ( 1705-1 772): silhouette (~1770) (Bach Museum, Eisenach) Thesonatasandpartitasforsoloviolinswv1001-1006 areamongJ.S.Bach'smostfamouscompositionsfor anymedium.Fortunately,theautograph faircopyof all sixworks-bearing thedate 1720-has survived; andnaturally,wearebestacquaintedwiththesetin the version found in this source.' But these pieces have also been handed down in a pre-1750source whichpresentsthecollectioninaconsiderablydiffer- entformat. Thesourceinquestionisamanuscriptinthehand oftheGrafenrodaKantor,JohannPeterKellner(1705- 1772), thatnowformsanindependentfascicleof the huge Bach miscellanyD-B Mus.ms.BachP804.*In- scriptions on the first page of the copy-'Scrips./ JohannPeterKellner/Anno1726JFrankenhayn.'-and itslastpage-'Frankenhayn. d.3. Jul:/l726.'-indicate not onlytheyear itwas made,but alsowhereitwas prepared and even the exact day on which it was completed. Thecopyisincomplete,omittingthewholePartita in B minor; the Allemande and Courante from the PartitainDminor;andtheLoure,Menuett11,Bourree, andGiguefromthePartitainEmajor.Furthermore,the works aregiveninaverydifferentorderfromthatin the autograph. Bach wrote out the set in the same familiarsequenceprinted intheBach- Werke-Verzeich- ni x3 G minorSonata,B minorPartita,AminorSonata, D minor Partita,C major Sonata,E major Partita;in Kellner's copytheorderis:Gminor Sonata,A minor Sonata, C major Sonata, E major Partita, D minor Partita. However,themostimportantdiscrepancybetween thetwomanuscriptsinvolvesvariantversionsofthree of thebest-known movementsin thecollection:the Chaconne fromthe D minor Partita and the fugues fromthe G minorandC major Sonatas.The Kellner copygives substantially shorterversions of allthree movementsandprovidesmaterialforthetwofugues not found inthe autograph. Thesevariants are also noteworthybecausetheyaretheonlyversionsof any of the movements from the collection, excluding transcription^,^thatdiffersignificantlyfromtheauto- graph.5Moreover,they differmore sharplythanany versions-transcriptions included-with regard to length. Our concernherewill be their authenticity. Inadditiontobeingtheonlysourceforthemodified versions, Kellner's copy is the only source which presentstheworksinthefragmentarystateordisjunct order outlined above. Evidently the source(s)from whichthecopywas derivedhavenot s ~r vi ved. ~ ScholarshavelongbeenawareofKellner'scopy,but eventhemostthoroughdiscussionsofthemanuscript arequitesuperficialandinconclusive.Forexample,in a recent essay Helmut Braunlichunhesitatingly ac- cepts all three variants as authentic early versions, EARLY MUSI C MAY 1 9 8 5 199 despite some fairly obvious clues to the ~ont r ar y. ~ fifth. This copy appears to derive from the same lost Further, in the critical commentary to NBA VI/i the only statement made regarding their authenticity is that they should in no way be interpreted as the results of 'unauthorized liberties' on Kellner's part, simply because of his 'deep respect for Bach.' Nowhere in either of these discussions is the manuscript com- pared to other Bach copies by Kellner, nor are there any remarks on his possible motives for preparing the source, two rather crucial issues in attempting to determine the authenticity of the Kellner versions. Perhaps the best way of approaching this topic is to consider Kellner's activities as a scribe and musician. He is without a doubt one of the most important copyists represented in Bach sources. Besides his personal acquaintance with Bach, Kellner was an extremely prolific copyist of Bach's compositions. In many instances Kellner's copy is the earliest-and in a few cases the only-surviving source for a Bach work. Despite his prominence in the transmission of Bach's music, Kellner is also unquestionably an ex- ceedingly careless scribe; this aspect of his work has received comment since the first half of the nineteenth ~e nt ur y. ~ This carelessness is plainly evident in his copies of the violin works, whose numerous errors range from incorrect notation of pitch and rhythm to inaccuracy of phrasing and the omission or duplication of whole bars. Of greater significance than Kellner's less than meticulous scribal methods is the fact that a few of his Bach copies transmit versions of works in which sizeable sections of material seem to have been arbitrarily excised. Three cases will suffice for example. Firstly, Kellner's copy of the Bach organ transcription of Vivaldi's D major violin concerto ~~208- t he so- called Grosso Mogul swv594-omits the lengthy cad- enza of its first movement entirely and gives a drastically abbreviated and apparently corrupt reading for its last movement cadenza.I0 Secondly, his copy of the Prelude and Fugue in C major ~ ~ ~ 5 3 1 (B Mus.ms. Bach P 274) presents a shortened version of the Fugue which must be considered corrupt. In addition to this, recent research has shown that the manuscript is derived from the same lost source as the copy in the Mollersche Handschrift (D-B Mus.ms.40644), which gives this Fugue in its standard form." Finally, Kellner's copy of the Aria variata swv989 (B ~us.ms.Bach P 804) omits the eighth and ninth variations as well as inserting the tenth variation between the fourth and manuscript as Johann Tobias Krebs' copy of the work (D-Bds Mus.ms.Bach P 801), which preserves the set in a much more conventional format.12 Since none of the exemplars for these three Kellner copies are extant, it is impossible to say with certainty just how responsible he was for these versions. But the conclusion that he modified them himself is suggested by the lack of any other sources except manuscripts prepared from his copies." Kellner's unreliability as a copyist aside, we should address the question of why he prepared his copy of the violin works. According to the autobiography of his son, Johann Christoph (1736-1 803), Kellner did play the violin, but to what degree of proficiency is unclear.14 Kellner was better known to his contem- poraries as an organ virtuoso; Johann Christoph's autobiography is the only source-contemporary or posthumous-which mentions his father in connect- ion with the violin. The only surviving instrumental compositions by Kellner are solo keyboard and organ piece^,'^ and he mentions no works for other instru- mental media in the work-list of his autobiography.16 Furthermore, only four of his Bach manuscripts, of which over forty survive, are copies of works other than solo keyboard and organ pieces." It should also be mentioned that four of Kellner's pupils, including Johann Christoph and Johann Philipp Kirnberger, are reported to have studied keyboard and organ with him, but violin with others1' (implying that Kellner lacked experience as a violinist?). Thus, it is obvious that Kellner's interest in the violin was subordinate to his activities in the realm of keyboard and organ music. It soon becomes a matter of considerable doubt that Kellner had the technical ability to have performed these works-among the most demanding in the violin repertoire-nor is there evidence that he might have prepared a copy from which someone else could perform. Assuming that Kellner copied out the sonatas and partitas for purposes other than violin performance, we should offer speculation as to other motives. While all that can be offered is unfortunately purely con- jectural, certain clues can be garnered from other Kellner copies as well as further details of biographical data. The most intriguing clue we have comes from Kellner's copy of another set of unaccompanied string compositions by Bach, the suites for solo cello 200 EARLY MUSI C MAY 1985 swv1007-1012 (B Mus.ms.Bach P 804). On the basis of its script, the copy of the cello suites, like that of the solo violin works, appears to date from c1726. It is complete, with the exception of the fifth suite (C minor) in which Bach calls for the a string to be tuned down a whole tone, and to judge from numerous pitch errors in his copy, Kellner's exemplar must have utilized this tuning. l9 Instead of retaining thescordatura in his copy, Kellner laboured to notate at sounding pitch all passages on the a string. Evidently the task proved to be so troublesome that the Sarabande was completely omitted and only the first nine bars of the Gigue were copied. It seems impossible that Kellner copied out the fifth suite for it to be performed on a cello. If one chooses to play the work without the scordatura, certain alter- ations have to be made in the chordal writing which Kellner did not incorporate into his copy.20 Keyboard or organ transcription suggests itself as a possible reason for the preparation of the copy. Kellner appears to have had a keen interest in this practice. He copied out a large number of Bach's keyboard transcriptions of concertos and more impor- tantly, there is evidence that he prepared organ arrangements of a lost Bach chamber work him~elf.~' It should also be pointed out that one of Kellner's students, Leonhard Frischmuth (c1700?-64), arranged six violin concertos by Tartini for harp~i chord, ~~ implying that Kellner may have used transcription as a pedagogical tool. In addition to Kellner's obvious interest in transcription, there is the fact that it is advantageous to work from a source which uses normal tuning when preparing a keyboard arrangement of a string composition. I am not proposing by any means, however, that Kellner planned on preparing keyboard and/or organ transcriptions of all six suites. The sheer size of such an undertaking would have made it prohibitive, coupled with the realization that newly-composed material would have had to have been added to provide harmonic support and contrapuntal interest if the arrangements were to have resembled the surviving keyboard and organ arrangements of the unaccom- panied violin pieces (see fn.4). It is far more plausible that Kellner wrote out the manuscript primarily to have a copy of the collection for reference purposes. Such study copies were common in the period. Whatever the motives, it seems quite clear that the fifth suite was not copied out with cello performance "LL&, >", . , 1 Violinist: engraving (Nurernberg. c 1720) in mind. No source mentions Kellner in connection with the cello, which, coupled with the sizeable technical demands posed by the set, makes it very doubtful that he wrote out the suites to be performed by himself or anyone else.23 The same is likely to be true of the violin works. The first of Kellner's modifications to be discussed-the Chaconne from the D minor Partita-presents more evidence to corroborate this theory. But first, a note on ordering in Kellner's copy of the violin works. When Bach altered the length of a composition in the process of revising it, he almost always made it 10nger.~' One might conjecture, &ere- fore, that Kellner's copy represents Bach's first draft of the collection, since three of its movements are considerably shorter than the versions in the auto- graph. Following this line of reasoning, one might also EARLY MUSI C MAY 1985 201 be tempted to speculate that Bach changed substan- tially the ordering of the set when he revised these movements because of the huge ordering discrepancies between the two sources. It is possible that Bach's original scheme was of two distinct halves-sonatas and partitas-the decision to alternate being taken in the revised plan. However, certain peculiarities in the ordering and numbering of the pieces in Kellner's copy suggest that his manuscript is derived from a source or sources which presented the works in the same order as found in the autograph. The ordering of the sonatas in Kellner's copy agrees entirely with the autograph; the ordering of the partitas, on the other hand, does not in any way agree. Though Bach uses the same numerical designations for the sonatas as the partitas: Sonata I ma , Partita I ma, Sonata 2 d 0 , Partita z d a , etc., Kellner uses different numerical designations for the two. He numbers the sonatas on the title page of his copy as well as in the individual work headings exactly as does Bach, where- as with the partitas no numbers are given in work headings, and those used on the title page appear after the respective titles: Partie in E#. I . and Partie in Db.2. These discrepancies imply that whoever arranged the works in the order found in Kellner's copy decided to retain numerical designations only for those pieces whose ordering agreed with the autograph's. From the previous discussion of Kellner's scribal methods, it is not beyond possibility that he chose to divide his copy into two distinct halves for whatever reasons, despite the numerical designations in his exemplars. The Chaconne from the D minor Partita in Kellner's manuscript (see illus. 1) is roughly three-fifths the size of the standard 257-bar version found in the autograph, omitting five passages that vary in length from four to forty bars (21-24,89-120,126-140,177-216 and 241- 244).25 The first (bars 21-24) and the last (bars 241- 244) of these omissions, each comprising an entire variation, can easily be explained as straightforward copying errors. But, because of the large amount of material lost and the type of passage-work entailed, the remaining three excluded sections do not appear to have been omitted through mere carelessness. The musical success of the modified form with so many bars missing is diminished by an unsatisfactory cadence resulting from the second omitted section (bars 89-120), and by the interruption of the Chacon- ne's regular 4-bar phrasing and an extremely abrupt modulation from G minor to D major arising from the 202 EARLY MUSI C MAY 1985 third (bars 126-40). The fourth and most substantial of the cuts (bars 177-2 16), on the other hand, does not show these weaknesses: there is a smooth and con- vincing transition between bars 176 and 217 in continuous semiquavers with a V-I cadence virtually identical to that at bars 244-5. Despite this, Kellner's copy is plagued by some major musical shortcomings which cannot be ignored. Another reason for questioning the authenticity of this version is the type of passagework in the three omitted segments just discussed, and it is in this regard that Kellner's motives for preparing his copy of the move- ment are pertinent. In an article published in 1920, the violinist and musicologist Andreas Moser proposed that Kellner excluded certain passages from his copy of the Chaconne because of the technical difficulties they pose for the ~i ol i ni st : ' ~ it is an argument not without basis. The three large segments missing from Kellner's copy unquestionably contain some of the most de- manding passages in the entire movement, most notably the famous arpeggiando material in bars 89- 120 and 201-208, but also the difficult quadruple stopping in bars 126-1 30. But we cannot presume that Kellner sought to produce a simplified performing score because his copy also preserves some of the most technically challenging passages of the Chaconne, including the demi-semiquaver figuration of bars 65-88. One might more successfully attempt to explain these omissions by arguing that Kellner prepared his copy of the Chaconne to transcribe it for keyboard or organ. Undoubtedly the two sections of the movement that pose the greatest problem in being transferred to a keyboard, being least idiomatic, are the two arpeggiando passages (unless transcribed as blocked chords). Fast scale passagework like that found in bars 65-88, on the other hand, is a hallmark of baroque keyboard writing. Further, the transcription hypothesis can in no way account for the omission of bars 126-140 and 177-200, passages which are easily adaptable to a keyboard. Both are characterized by extensive multiple- stop writing, a feature common throughout Kellner's COPY. Faced with this puzzling state of affairs, yet another possibility should be offered: that Kellner intended his copy of the sonatas and partitas to serve as nothing more than a 'study copy', perhaps complementing that of the cello suites, deleting certain passages from the Chaconne simply because of its length. This brief excursion must conclude with the admis- I Kellner's copy of the Chaconne Rom Bach's Partita in D minor BWV1005. D-B - P 804. fasc.22 f 12" sion that it is by no means certain that Kellner himself is responsible for these omissions; the scribe of Kellner's exemplar may have excluded material for whatever reasons, Kellner thereafter preparing what he thought to be an unabridged copy of the movement. The evidence would, however, lead us to suspect that, for one reason or another, Kellner knowingly excised portions of the Chaconne as he copied it. Kellner's possible motives for abbreviating the Cha- conne are of course not nearly as important as the realization that this variant, obviously conupt, cannot be authentic Bach. The two variants that remain to be discussed, however-abbreviated versions of the fugues from the G minor and C major sonatas-do appear to represent authentic early versions of these movements. EARLY MUSI C MAY 1985 203 I (contd.) Kellner's copy of the Chaconne (f.13r) I # I . J . .. . . - 4 -*- - + -..*.! $*&!. ,!- - - . - -.. , . . - .. - Kellnefs copy of the Fugue from the G minor Sonata (illus.2) omits only a single 7-bar segment of the stand- ard autograph version: bars 35-41. The multiple stopping in this passage makes it one of the most problematic in the movement to perform, substanti- ating at first glance Moser's view that Kellner excluded material from the Chaconne because of its technical difficulty. If one accepts David Boyden's theory that, beginning with the third beat of bar 35, this passage was probably meant to be arpeggiated because of its chordal nature, a further parallel to the Chaconne copy would appear to exist.27 Unlike the copy of the Chaconne, however, this variant gives material not found in the autograph version in the bars immediately adjacent to the omitted passage (ex.1) which in no way raises doubt 204 EARLY MUSIC MAY 1985 I (contd.) Kellner's copy of the Chaconne I - -- (t 13v) / ' 0 pqW ' < . d . - < r --e,,..f,l 4 w r .-t ,J-- $, Ex.] Kehel s modified reading from the Fugue. Sonata no.1 movement-especially in the upper two voices of bars (G minor) BWIOOI 58-62 of the autograph version-and they lead to a perfectly effective cadence on the downbeat of bar 35. Still more significant are the striking similarities between the latter half of bar 34 from Kellner's copy and the last two beats of the upper voice in bar 63 from about its authenticity. The parallel sixths in bars 34- the autograph. Both of these cadential figures share 35 of Kellner's copy are in abundance throughout the the same rhythm-two quavers, four semiquavers- EARLY MUSI C MAY 1985 205 2 Kellner's copy of the Fugue from Bach's Sonata in C minor BwV1001. D-B P 804. fasc.22 f.2v and general melodic contour. The two quavers in both downward descent. It is also worth noting that both involve a downward leap: in Kellner's copy a sixth, in passages immediately precede episodes which are bar 63 of the autograph a perfect fourth. Moreover, the exactly ten bars long and comprise semiquavers second quaver in each passage is immediately followed exclusively. by a semiquaver a second lower. Finally, the semi- The only major difficulty in accepting the authen- quavers in both lead to a lower pitch, concluding the ticity of this variant stems from our thorough familiarity 206 EARLY MUSI C MAY 1985 2 (contd.) Kellner's copy of the C minor Fugue (f.31) with the autograph version. We are so accustomed to making the passage surely one of the most unforgett- hearing bars 35-41 that we inevitably miss them upon able in the movement. But this, obviously, does not hearing a version of the fugue in which they are constitute sufficient reason to question the authen- absent. This is particularly true of bars 38-41, where a ticity of the variant. series of quavers in parallel thirds and sixths is introduced over a tension-generating pedal point. Kellner's copy of the fugue from the C major Sonata EARLY MUSIC MAY 1985 207 (illus.3) while also evidently representing an authentic early version, is far from an accurate transmission of such a version. It omits bars 188-200, 256-270 and 277-286, but in each instance presents material not found in the autograph version, just as in the variant of the G minor fugue. The excluded segments are all clearly derived from material included in the variant and are not among the movement's most technically challenging passages. The passage beginning with bar 186 in Kellner's copy (see ex.2) is plainly corrupt since it provides only two beats for bar 187. The subsequent bar, which leads directly to one identical to bar 201, does not appear in the autograph. Evidently, Kellner or the scribe of his exemplar glanced from the second beat of bar 187 to the downbeat of this bar, its first two beats being identical to the last two of bar 187. Ex.2 Kellner's modified reading from the Fugue, Sonata no.3 (C major) BW1005 [bar 18.91 ILU u [autograph bar 2OlJ A later passage from Kellner's copy which corre- sponds to ex.2 is a good clue as to the amount of missing material. (See ex.3, beginning with the notes beneath the asterisk.) In ex.3 this passage is five bars long, with the fugue subject stated once, followed by cadential material. The last beat of the bar corresponding to bar 276 and the downbeat of that corresponding to bar 287 differ from the autograph. Although the cadence in ex.3 sounds to us premature, because of our familiarity with the autograph version, there is nothing in the reading which might lead us to question its authen- ticity. It would seem then that Bach expanded this passage by ten bars when he prepared the autograph version. In the autograph the passages which correspond to exx.2 and 3 are both exactly fifteen bars long, in addition to using virtually the same material. It is not unreasonable to assume that they were the same length in the early version too and that two and a half bars are missing from ex.2. A few observations on the first system of ex.3 should also be offered. It begins with two bars identical to 254-255 and then proceeds directly to a bar not found 208 EARLY MUSI C MAY 1985 Ex.3 Kellner's modified reading from the Fugue, Sonata 110.3 (C major) ~ ~ 1 0 0 5 [autograph bar 2541 [autograph bar 2711 [autograph bar 2761 , - [autograph bar 2.971 1 1 , I in the autograph, which, in turn, leads to a bar identical to 27 1. The bar unique to Kellner's copy uses a fi gur~whkh appears five times in the autograph (see the f fJ'r figure in bars 43, 79, 81, 83 and 331), and in each case, as here, is immediately preceded by four quavers or a motive. Considering the uncontrived manner of the passage, plus its motivic derivation, one is hard-pressed to point to any musical shortcomings of the reading. One feature of ex.3, however, does arouse suspicion. Beginning on the third beat of bar 288, Bach repeats the opening bars of the movement as far as the downbeat of bar 65, and in his fair copy he writes out all the notes of the repeat, instead of using a da capo indication. But Kellner writes out only the first minim of the restatement and in the bar corresponding to 289 provides a 'Da Capo' inscription. If the fugue involved a literal repeat from bar 289 until the end, Kellner's da capo indication would barely matter, but bars 289-296 are in no way a literal restatement of 1-8. Bach adds counterpoint to the first two statements of the subject, beginning with the third beat of bar 289 and extending to the second beat of 296, from which point until bar 353 the repeat is literal; only the final chord differs. This discrepancy between Kellner's copy and the autograph led Helmut Braunlich to maintain that Bach originally conceived this movement as containing a literal da capo, and that he added counterpoint to the first several bars of the restatement only when he prepared the autograph version.** While this theory seems logical enough on the surface, a more careful examination of Kellner's copy leads to a different conclusion. 5 Kellner's copy of the Fugue from Bach's Sonata in C major BwV1005. D-B P 804. fasc.22ff 6v-81 EARLY MUSIC MAY 1985 209 In the custos immediately prior to his 'Da Capo' inscription, Kellner shows that the repeat begins on a'; yet there is no segno on the downbeat of bar 1, nor is there a fermata on the downbeat of bar 66, indicating the conclusion of the movement. But both of these omissions, suspicious as they are, could be explained as copying oversights. The minim rest which appears directly beneath the custos, though, strongly suggests that the da capo restatement of the early version was identical to that in the autograph because the rest clearly belongs only in bar 289: neither in Kellner's copy nor in the autograph does a rest appear in bar 1. It is also significant that in Kellner's copy the rest is obviously intended for the lower voice, occupying the lowest space of the system, as it does in bar 289 of the autograph. Evidently, Kellner (or possibly the scribe of his exemplar) realized that the movement was a da capo fugue immediately after notating the rest. Following this line of reasoning, the scribe guilty of the omission must have also realized that the repeat was not literal, but decided to spare himself the trouble of writing it out, since only a short passage differed. What is puzzling is why he did not copy out the repeat until the third beat of the bar corresponding to 296 and then provide the da capo indication for the material which is a literal restatement. Lack of space was no problem in Kellner's case because his 'Da Capo' inscription appears on the top system of a page. Copying music is a more or less mechanical process, and though I may be assuming more discernment on the part of a scribe than is normally done, there seems little other expla- nation for the minim rest. To judge from the surviving evidence, Bach revised the C major and G minor Fugues by interpolating passages which share common features. Bars 38-41 of the G minor Fugue and 188-200 and 277-286 of the C major Fugue play a crucial structural role in these movements and they achieve this status thqough very similar means. All three passages involve extensive pedal points that lead to major cadences, cadences which in each instance are dramatically delayed by repetition or sequence in the upper voices. It seems that Bach also changed the metre of both movements from C to $.29 These similarities imply that the Fugues may have been revised at around the same time. Regrettably, it is impossible to determine when these revisions might have been made. The autograph-a fair copy rather than a composing score-supplies only a terminus ante quem of 1720. 210 EARLY MUSI C MAY 1985 Johann Peter Kellner's copy of Bach's sonatas and partitas for solo violin not only preserves what appear to be early versions of two movements from the collection. It also reminds us how cautious we ought to be in gauging the value of secondary sources and how many various aspects of these sources we should weigh in measuring their worth. It is hoped that this essay has demonstrated how necessary and potentially valuable both considerations are. This study is a revised version of a paper delivered at the 1983fall meeting of the midwest chapter of the American Musicological Society (Chicago). I would like to thank Professor Robert L. Marshall (Brandeis University) for many helpful comments and suggestions. Russell Stinson is a doctoral candidate in musicology at the University of Chicago, where he is completing a dissertation on the Bach manuscripts of Johann Peter Kellner and his circle. ID-B Mus.ms.Bach P 967. A facsimile of the manuscript can be found in G. Hausswald, ed., Johonn Sebastian Bach: Sonaten und Partiten fiir Violine allein (Frankfurt am Main, 1962/R1982, with Eng. trans. of foreword). % complete inventory of the manuscript can be found in NBA V/V, KB, pp.24-35. 3WW. Schmieder, Thematisch-systematischesVerzeichnis der rnusihal- ischen Werhe von Johann Sebastian Bach (Leipzig, 1950), pp.559-562 Vhese include keyboard arrangements of the A minor Sonata BWV964, and the first movement of the C major Sonata BWv968; what is most probably a transcription for lute of the E major Partita Bwv1006a; an arrangement of the Preludio from the E major Partita for orchestra with organ obbligato swv29/1 and 1200/1; and organ and lute transcriptions of the Fugue from the G minor Sonata ~ ~ ~ 5 3 9 1 2 and 1000, respectively. Only in the last instance is there any discrepancy in length of the movement, with both arrangements being two bars longer than the violin version. On the question of medium in BWv1006a see NBA V/x, KB, pp.167-170. 5For detailed listings of discrepancies among the sources for the set, see NBA VI/i, KB, pp.35-117. 6According to Hausswald and Gerber, the fragmentary state of Kellner's copy as well as its ordering shows that it was prepared from copies of individual works rather than a copy of the entire collection; see NBA VI/i, KB, pp.34f. 'H. Braunlich, 'Johann Peter Kellner's Copy of the Sonatas and Partitas for Violin Solo by J. S. Bach', Bach [The Quarterly Journal of the Riemenschneider Bach Institute], xii (1981), no.2, pp.2-10 8NBA, VI/i, KB, pp.28f 9Friedrich Konrad Griepenkerl(l782-1849), one of the co-editors of the Peters edition of Bach's complete organ works, is reported to have referred to Kellner as 'ein sehr nachlassiger Abschreiber'; see BG, XLVf, p.lv. 'Karl-Marx-Universitat, Leipzig, Inv.5137; cf. NBA IV/viii, KB, pp.44f and 49f. "Kilian tentatively assigns B Mus.ms.Bach P 274 to Wolfgang Nicolaus Mey, a copyist whose script is remarkably similar to the last two phases of Kellner's handwriting: watermark as well as handwriting evidence reveals that Kellner was the scribe, See NBA IV/v-vi, KB, pp.284-285 [for another view see G. Stauffer's article, this issue, p. 1951. I2The copy. D-Bds Mus.ms.Bach P 801 also omits variation 9, but gives the variations in the same sequence as all the other sources for the work; see NBA V/x, KB, pp.43-47. Plath (NBA Vlv, KB, p.29) and Kilian (NBA IV/v-vi, KB, p.196) question whether Kellner is the scribe of this copy too; but again it appears certain that Kellner prepared this source, a conclusion which Eichberg accepts (NBA V/x, KB, p.41). '3Kellner's copies of swv53 1 and 594 served as exemplars for D-B Mus,ms.Bach P 286: see NBA IV/v-vi, KB, pp.285f [BWV531] and NBA IV/viii, KB, p.46 [Bwv594]. Kellner is reported to have prepared a copy of the Prelude and Fugue in G major BWV550, which omitted many bars from the conclusion of the Prelude, so this lost manuscript may represent yet another example of a corrupt abbreviated version. Peter Williams offers this as a possibility in The Organ Music of J. S. Bach, i (Cambridge, 1980), p.176. Kellner's lost copy of swv550 appears to have been the exemplar for copies in D-B Mus.ms.Bach P 642 and 924: see NBA IV/v-vi, KB, pp.421f. I4Johann Christoph makes only the following statement about his father's violin playing, apropos of his own desire to learn the instrument: 'Mein Vater, der den Nutzen der praktischen Kenntnisse dieses Instruments fiir einen kunstigen Tonsetzer kannte, stimmte sehr gern mit meiner Neigung uberein' (p.43). See F. W. Strieder, Grundlage z u einer Hessischen Gelehrten und SchrifsteNer Geschichte, vii (Kassel, 1787), pp.41-8; an excerpt is published in B-Dok iii, p.435. I5Kellner also composed numerous church cantatas, none of them published; many survive in the Stadt- und Universitats- bibliothek Frankfurt am Main. 16Kellner's autobiography is in Marpurg's Historische- kritische Beytriige zur Aufnahme der Musik, i (Berlin, 1754/R1970), pp.439ff. See B-Dok iii, p.77, for an excerpt. "The copy of the violin works and copies of the six cello suites ~ ~ ~ 1 0 0 7 - 1 0 the Prelude in C 12, the E minor flute Sonata ~ ~ ~ 1 0 3 4 , minor for lute BwV999, all found in B Mus.ms.Bach P 804. Kellner was also the scribe for the soloist's part, and the beginning and end of the first violin part of the D minor Harpsichord Concerto swv1052 (B Mus.ms.Bach St 125). I8In Johann Christoph's case this is clear from his autobiography; concerning Kirnberger, see B-Dok iii, pp.75f. The other two students were Johann Valentin Scherlitz (1732-93) and Johann Georg Gressler (1732-?). On Scherlitz, see E. L. Gerber, Neues historisch- biographisches Lexikon der Tonkunstler, iv (Leipzig, 1814/R1966), pp.52f; on GresSler, see J. G. Bruckner, Sammlung verschiedener Nachrichten z u einer Beschreibung des Kirchen- und Schulenstaates im Herzogthum Gotha, iii (Gotha, 1761), pp.70f. 19This is noted bv Alfred Dorffel in his edition of the cello suites for the BG, xxvii, p.xxxlv. ZoHugo Becker, in his International Music Company edition of the cello suites (New York, 1946). includes a version of the fifth suite with normal tuning which clearly shows these necessary alterations. "These transcriptions, which survive in B Mus.ms.Bach P 804, fasc.12, and P 288, fasc.4, will be discussed at length in my dissertation on the Kellner-circle Bach manuscripts (U, of Chicago). "M. Dounias, Die Violinkonzerte Giuseppe Tartinis (Wolfenbuttel, 1935), p.198 231 should also mention here that nowhere in Kellner's copy of the sixth cello suite (D major) is there any indication that the work is intended for an instrument with five strings, implying that he copied out the cello suites for reasons other than cello performance. 240ne exception is the Prelude in E minor ~WV53311, whose early version is two bars longer than its revised form. See NBA IV/v-vi, KB, p.382f. ZVor the sake of convenience, the bar numbers here correspond to those in NBA VI/i, even though the first two beats of the movement are counted as bar one there. %ee A. Moser, 'Zu Joh. Seb. Bachs Sonaten und Partiten fur Violine allein', BJb, xvii (1920), pp.30-65; esp. p.35 fn.1. "D. Boyden, The History of Violin Playing from its Origins to 1760 (London, 1965), p.439, fn.19. Had Kellner realized that bars 35-41 were meant to be played arpeggiando-perhaps because of an 'arpeggio' inscription in his exemplar-it is not too difficult to suppose that he could have omitted the passage and inserted his own material to bridge from bars 34 to 42. But it appears doubtful that Bach intended this passage to be arpeggiated since he does not indicate any kind of arpeggiation in the autograph (nor is arpeg- giation prescribed in any of the other sources for the movement). It seems rather unlikely that Bach would have omitted such an important indication-in what is undoubtedly one of his most meticulously prepared holographs-when he provides such detailed indications for the arpeggiando passages from the Chaconne. 28Braunlich,op cit, p.4 29While Kellner's copy gives C for both Fugues, the autograph gives 4. EARLY MUSI C MAY 1985 21 1