You are on page 1of 27

The Printing History of J.S.

Bach's Musical Offering: New Interpretations Author(s): Gregory Butler Reviewed work(s): Source: The Journal of Musicology, Vol. 19, No. 2 (Spring 2002), pp. 306-331 Published by: University of California Press Stable URL: http://www.jstor.org/stable/10.1525/jm.2002.19.2.306 . Accessed: 08/11/2012 12:33
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

University of California Press is collaborating with JSTOR to digitize, preserve and extend access to The Journal of Musicology.

http://www.jstor.org

The Printing History of J. S. Bachs Musical Offering: New Interpretations


GREGOR Y BUTLER

306

n his monumental Bach biography, Philipp Spitta pointed to a piecemeal production of the various printing units that comprise the original print of Bachs Musical Offering and suggested that no particular disposition scheme was ever intended by the composer for the collection as a whole.1 His installment theory (Lieferungstheorie) went largely unchallenged for more than half a century until Hans David took issue with Spitta, arguing that the collection was conceived by Bach, engraved and printed, and dispatched to Friedrich II of Prussia as an integral whole.2 Davids emphasis on the integral, cyclical nature of the collection subsequently gave rise to a number of studies focusing on the question of ordering.3 During this period only
Volume XIX Number 2 Spring 2002 The Journal of Musicology 2002 by the Regents of the University of California
A section of this essay was presented at the International Bach Symposium in Leipzig, Bach in Leipzig: Bach und Leipzig, on 28 January 2000, and appeared as Eine neue Interpretation der Druckgeschichte des Musikalischen Opfers in Bach in Leipzig Bach und Leipzig. Konferenzbericht Leipzig 2000, edited by Ulrich Leisinger (Hildesheim: Georg Olms, 2002), 30920. I am grateful to those scholars who have so kindly read this study and commented on it: Michael Marissen, Joshua Rifkin, Christoph Wolff, and Neal Zaslaw. I would like to acknowledge the Social Sciences and Humanities Research Council of Canada for their generous support.
1 Philipp Spitta, Johann Sebastian Bach, 2 vols. (Leipzig: Breitkopf & Hrtel, 1873 80), 2:67176, 84345; English trans. Clara Bell and J. A. Fuller-Maitland, 3 vols. (London: Novello, 1889), 3:19197, 29294. 2 Hans David, Bachs Musical Offering, Musical Quarterly 23 (1937): 31432. 3 Hans David, J. S. Bachs Musical Offering. History, Interpretation, and Analysis (New York: Dover Publications, 1945); Rudolf Gerber, Sinn und Ordnung in Bachs Musikalischem Opfer, Das Musikleben 1 (1948): 65ff.; and Wilhelm Pfannkuch, Das Musikalische Opfer, in Die Werke des Bach-Festes 1950 (Kassel: Brenreiter, 1950), 20ff.

butler
one scholar, Friedrich Blume, supported Spittas Lieferungstheorie, re ning it by introducing the idea that the collection evolved from an original nucleus by a process of progressive expansion.4 Then in the early 1970s Christoph Wolff, on the basis of his rigorous diplomatic examination of the original print,5 sought to alter the prevailing view of the collection as an integral cycle.6 Nonetheless, the earlier emphasis on ordering seems once again to have taken center stage. Ursula Kirkendales controversial application of the disposition scheme of classical forensic rhetoric to the work to arrive at a de nitive ordering of its component parts has, at least on this issue, divided scholars into two camps.7 More recently, Michael Marissen has argued for an ordering based on references to the work in secondary sources in conjunction with evidence provided by the original print and the surviving autograph materials.8 In this study my goal is to arrive at a clearer understanding of the printing history of the collection by undertaking a detailed reexamination of all surviving exemplars of the various printing units.

The Installment Theory


Spitta was the rst scholar to identify A 1,9 one of the partial exemplars in the Amalienbibliothek, Berlin as the dedicatory copy of the work sent to the king. A long entry in the appendix to the second volume of his Bach biography is devoted to what for the time is a detailed diplomatic description of this source:
The dedicatory exemplar contains: 1) three leaves with music and two leaves with title and dedication. The paper is of unusual beauty and weight; the format, largest oblong folio. These leaves are in leather boards with gold tooling. The musical contents comprise the threevoice simple fugue and a canon in which the alto presents the cantus
Friedrich Blume, Umrisse eines neuen Bach-Bildes, Musica 16 (1962): 176. See J. S. Bach. Kanons, Musikalisches Opfer, ed. Christoph Wolff, Neue Bach-Ausgabe smtlicher Werke, series 8, vol. 1. Kritischer Bericht (Kassel: Brenreiter, 1976), hereafter, NBA 8/1 KB. 6 Christoph Wolff, New Research on Bachs Musical Offering, Musical Quarterly 57 (1971): 379408. A revised version under the same title appears as chapter 18 in Christoph Wolff, Bach: Essays on His Life and Music (Cambridge: Harvard Univ. Press, 1991), 23958. All page references are to this later, revised edition. For his view that Bach never intended a particular ordering for a complete cyclical performance, see especially 258. 7 Ursula Kirkendale, The Source for Bachs Musical Offering: The Istitutio Oratoria of Quintilian, Journal of the American Musicological Society 33 (1980): 88141. Wolff s critique of Kirkendales interpretation appears as a postscript to Wolff, New Research that appears after the endnotes, 42123. 8 Michael Marissen, More Source-Critical Research on Bachs Musical Offering, Bach 25 (1994): 1127. 9 I have retained the numbering of the various exemplars of the original print established by Wolff in the critical notes to NBA 8/1 KB. For a detailed diplomatic description of this exemplar, see 5862.
4 5

307

t he j ou r na l o f mus ic o log y
rmus while discant and bass proceed in counterpoint canonically. The canon has the engraved title Canon perpetuus super Thema Regium; the fugue is called Ricercar. 2) an upright folio leaf of the same physical appearance with regard to dimensions and quality of the paper. Because of its divergent format, it is only a loose enclosure, and contains on the two inner pages under the engraved title Canones diversi super Thema Regium, ve canons and a Fuga canonica in Epidiapente. The leaves in oblong folio as well as those in upright folio carry written inscriptions as well. Besides the compliments [to the king] appended to the fourth and fth canons of the upright folio leaf on the right side (discussed by me in the text), on the rst blank page of the same appears the title Thematis Regii Elaborationes Canonicae; on the rst blank page of the leaves of music in oblong folio as well: Regis Iussu Cantio Et Reliqua Canonica Arte Resoluta.10

308

At the time of writing, according to Spitta, his two units1) the title page and dedication, and the three-part ricercar with its appended canon (printing units A and B)11 and 2) the diverse canons (printing unit D)formed a part of a complete exemplar of the original print (only the title wrapper to the sonata was missing). Because the remaining two printing units from this exemplarthe six-part ricercar with its two appended canons (printing unit E) whose four leaves are presently enclosed loose as Beilage 2 in exemplar A 1 and the sonata with its appended single canon (printing unit C) which is catalogued separately 12 appeared to be printed on the same paper as the exemplars from the

10 Spitta, Johann Sebastian Bach, 2:843: Das Dedications-Exemplar enthlt: 1) 3 Bltter mit Noten und zwei Bltter mit Titel und Widmung. Das Papier ist von seltener Schnheit und Strke, das Format grtes Querfolio. Diese fnf Bltter sind in braunem Lederband mit Goldpressung. Den musikalischen Inhalt bilden die dreistimmige einfache Fuge und ein Canon, in welchem der Alt den Cantus rmus fhrt, whrend Discant und Bass canonisch contrapunktiren. Der Canon hat die gestochene berschrift Canon perpetuus super Thema Regium, die Fuge ist Ricercar benannt. 2) Einen Hochfolio-Bogen von derselben Beschaffenheit in Bezug auf Gre und Gte des Papiers. Er ist seines abweichenden Formates wegen nur beigelegt, und enthlt auf den beiden Innenseiten unter der gestochenen berschrift Canones diversi super Thema Regium fnf Canons und eine Fuga canonica in Epidiapente. Die Bltter in Querfolio sowohl wie in Hochfolio tragen aber geschriebene Zustze. Auer den Glckwnschen, welche dem 4. und 5. Canon des Hochfoliobogens rechts zur Seite beigegeben und von mir im Contexte mitgetheilt sind, ndet sich auf der ersten leeren Seite desselben der Titel: Thematis Regii elaborationes canonicae; auf der ersten leeren Seite der Notenbltter in Querfolio aber: Regis Jussu Cantio Et Reliqua Canonica Arte Resoluta. Unless otherwise indicated, all translations from the German are mine. 11 Throughout this study I use Wolff s designation by capital letters in bold typeface for the ve printing units comprising the collection: A (title and dedication), B (threepart ricercar and perpetual canon), C (trio sonata and perpetual canon), D (various canons and canonic fugue), E (six-part ricercar and remaining two canons). Wolff, NBA 8/1 KB, 4849. 12 This is exemplar A 2. Ibid., 6263.

butler
main printing run he had examined,13 he concluded that only those three printing units printed on the oversized luxurious paper (A, B, and D) had been sent to the dedicatee originally while the other two printing unitsC and Ewere composed by Bach later and subsequently sent to the king without any particular disposition.14 This is the installment theory (Lieferungstheorie). In Chapter 7 of his book on the Musical Offering, in a section entitled Erroneous Conclusions,15 David challenges Spittas Lieferungstheorie. David claims that the Ricercar 6 Spitta thought formed a second installment, and the Sonata a third. But Spitta had stated only that the remaining portion of the work because of its outward form [i.e., different paper formats] divides into two parts and moreover that Bach had followed up his rst gift with a second of more importance, containing the six-part Ricercar, the Sonata, and three more canons. Spitta described the binding of the oblong sheets in section 1 of his description but nowhere does he state that it was made for Bach as David claims. Nor had Spitta argued, as David suggests, that the existence of sections in both oblong and upright format was due to a delay in the manufacture of the edition. Finally, the scenario proposed by Spitta, according to which the luxurious paper used for the printing of sections 1 and 2 had become unavailable at the time the rest of the collection was ready for printing is by no means a none too probable coincidence as David refers to it. On the contrary, as will be demonstrated presently, it falls well within the realm of possibility given the likely timetable for the production of the print. Believing he had demolished Spittas Lieferungstheorie, David states his belief that Bach devised and dispatched the Musical Offering as a whole. Davids conclusions, founded as they are on a misinterpretation of Spittas position, are clearly open to question. Nevertheless, they have had a strong impact on all research on the Musical Offering since their publication. For instance, Wolff, in the critical report accompanying his edition of the work for the Neue Bach-Ausgabe, is of the same opinion as David, believing that Bachs dedicatory copy was originally a complete exemplar printed in its entirety on the luxurious oversized paper on which Spittas sections 1 and 2 are printed. He argues: It would have been incomprehensible for Bach to have presented the Musical Offering, one part in ceremonial garb and the other part in its normal guise. Thus the dedicatory exemplar must be considered as
13 One of the three exemplars of the original print held by the sterreichische Nationalbibliothek, A 9 (recently recatalogued as A:Wn S.H. J.S.Bach 100), was presented to Spitta by Wilhelm Rust on 1 February 1875. Ibid., 65. 14 Spitta, Johann Sebastian Bach, 2:844: . . . ist von Bach spter componirt und nachtrglich ohne besondere Frmlichkeiten dem Knige zugesandt. 15 David, J. S. Bachs Musical Offering, 9496.

309

t he j ou r na l o f mus ic o log y
preserved only in a fragmentary state.16 As a consequence he concludes that printing unit E in exemplar A 1 and printing unit C in exemplar A 2 did not originally belong to the dedication copy but apparently were later replacements for the lost originals. These must also have been printed on luxurious paper like the present sections A, B, and D but they were somehow lost, very probably before the dedication copy was given to the library of Princess Amalia.17 This rejection of Spittas Lieferungstheorie is a linchpin in Wolffs wrapper theory which holds that all ve printing units were printed and published together and that for all exemplars of the main printing run, printing units A and D, which both came off the press as bifolios, acted as wrappers for printing units B and E respectively, an arrangement that gives rise to three distinct fascicles.18

The Engraving
For some time there has been no question of the identity of the engravers of the collection, the Schbler brothers from Zella in the Thuringian forest. The plates for printing units B and E were engraved by Johann Georg Schbler19 whereas those for printing unit D were the work of his younger brother, Johann Heinrich Schbler.20 It seems to have escaped notice that there was a third person involved in the engraving of the plates for printing unit C, the parts for the sonata. Johann Heinrich, the youngest and least skilled of the Schbler brothers was responsible for only a single plate, that for page 2 of the Flauto part. His older brother, Johann Georg, was the engraver of all four plates for the continuo part and pages 2 and 3 of the Violino part. The third engraver, who prepared the plates for pages 1, 3 and 4 of the Flauto part and pages 1 and 4 of the Violino part was clearly the most skilled; his cuts are bold and assured. His engraving differs from that of Johann Georg chie y in the form of the larger at sign with its very wide lower loop and the downward slant of the upper horizontal stroke of his eighth rests, but there are other more subtle differences as well. It is clear that Johann Heinrich was modeling his engraving after that

310

16 Wolff, NBA 8/1 KB, 60: Es wre unverstndlich, wenn Bach das Musikalische Opfer zu einem Teil in Prunkausstattung, zum anderen Teil in gewhnlicher Aufmachung berreicht htte. So mu das Widmungsexemplar als nur fragmentarisch erhalten gelten. 17 Wolff, New Research, 251. 18 Ibid., 24851. 19 The bottom right hand corner of the plate for the last page of printing unit E carries the inscription J. G. Schbler sc. 20 Wolfgang Wiemer discovered that a second engraver was involved in the project and identi ed him through concordances as the younger brother of Johann Georg, Johann Heinrich Schbler. Wolfgang Wiemer, Johann Heinrich Schbler, der Stecher der Kunst der Fuge, Bach-Jahrbuch 65 (1979): 77.

butler
of this more highly skilled engraver. Figure 1 offers examples of the work of the three engravers. Johann Heinrich, in his biographical sketch of about 1801/2, writes of his instruction in music and engraving as follows:
In his 10th year he was led by his brothers to music and to drawing. After his school years were nished he was instructed in engraving and iron cutting by his eldest brother, in music by his second brother who studied music in Leipzig with the famed Bach, and in gunstock carving by his father.21

His eldest brother, actually Johann Heinrichs half-brother, was Johann Jacob Friedrich Schbler, who was 24 years old in 1747. The youngest brother, Johann Heinrich, was initiated into the craft of engraving not by the second brother, Johann Georg, but by the eldest brother, and it is clear from the style of his engraving that Johann Georg also received his early instruction in engraving from Johann Jacob Friedrich. The eldest brother thus emerges as the master engraver of the clan. I would submit that Johann Jacob Friedrich was the third, most skilled of the three engravers involved in the engraving of the plates for printing unit C.

311

The Paper
Paper y,22 the oversized luxurious paper on which the dedicatory copy was pulled, according to Wolff was available only in single sheets and not in bifolios. Thus there was no possibility for forming three fascicles with separate title covers as in the ordinary copies.23 Because paper y is without watermark and because printing units A, B, and D of the dedicatory copy in its present state consist of a series of single leaves, it would seem to be impossible to ascertain the original dimensions and format of this paper. It is for this reason that Wolff states: Now whereas the various printing units of the regular exemplars of the original print consist alternately of bifolios and single leaves, this exemplar (the dedicatory copy), before the intervention of the bookbinder, existed only as single leaves.24 But Spitta refers three times in the course of his diplomatic
21 Ibid., 79: . . . in seinem 10.ten Jahre wurde er schon von seinen Brdern zur Music und zur Zeichnung angefhret. Nach geendigten Schuljahren wurde er von seinem ltesten Bruder in Graviren u: Eisenschneiden, von seinem zweyten Bruder, welcher die Music in Leipzig bey dem berhmten Bach gelernet, in der Music u: von seinem Vater in der Schfters Profession unterrichtet. 22 My lettering of the papers corresponds to that adopted by Wolff. See Wolff, NBA 8/1 KB, 5152. 23 Wolff, New Research, 251. 24 Wolff, NBA 8/1 KB, 58: Whrend nun die verschiedenen Druckeinheiten der normalen Exemplare des Originaldruckes wechselweise aus Bogen und Einzelblttern bestehen, hat dieses Exemplar vor dem buchbinderischen Eingriff nur aus Einzelblttern bestanden.

t he j ou r na l o f mus ic o log y
gure 1. Samples of engraving from the original print of J. S. Bachs Musical Offering by a) Johann Heinrich Schbler, b) Johann Georg Schbler and c) Johann Jacob Friedrich Schbler. 1977 by C. F. Peters Corporation. Used by permission. All rights reserved.

312

description to his section 2 (printing unit D) as an upright folio leaf (Hochfolio-Bogen), that is, a single leaf folded in two at the midpoint of the long side to produce a bifolio in upright format. Furthermore, he refers to it in the singular and speaks of its inner pages (Innenseiten). At some point after Spitta examined this source in connection with the publication of the second volume of his Bach biography in 1880, this folio leaf must have been cut along the fold to produce the two single leaves in existence today, almost certainly in the process of binding.

butler
This would seem to be at odds with Wolffs assertion that paper y was available only in single sheets and not in bifolios. The most problematic aspect of his argument is that even allowing for extensive cutting in the process of binding, the dimensions of the single sheets of paper y, 28.8 3 45.1 cm., correspond to none of the standard paper sizes available at the time the collection was printed. But as Spittas description suggests, this paper did originally exist as a large leaf having twice the dimensions of the single leaves and folded at the mid-line of the long side to produce a bifolio. The full sheets of paper y in their uncut state would have measured not less than 45.1 3 57.6 cm. Because it is clear that they were cut in the process of binding, both dimensions would have been somewhat greater originally. In fact, a paper with such dimensions did exist in Bachs day. This was the largest size of paper commonly available at that time, Royalpapier, a deluxe Doppelpapier whose sheets measured 48 3 64 cm.25 giving single sheet dimensions of 48 3 32 cm., approximately 3 cm. greater than either dimension of paper y in its present state. Thus Wolff s valid conclusion that printing units A and D printed on paper v came off the press as uncut folio leaves in the main printing run is equally valid for those same printing units in the dedicatory exemplar printed on paper y. Let us continue with our detailed study of the various papers used for the printing of the collection in an attempt to arrive at a clearer picture of the printing history of the Musical Offering. Whereas paper y is without watermark, all of the three papers employed in the main printing are identi able. Paper v 26 is a paper from the paper mill of Ephraim Lenk at Niederloessnitz in Saxony. This and other papers of Lenk, along with papers of the Vodel family of papermakers from Niederlungwitz, were widely used in Leipzig and were commonly found in Bernhard Christoph Breitkopf s prints throughout the years of Bachs tenure in Leipzig.27 Printing unit A (title page and dedication), the only unit set in movable type, was pulled on paper v in one of Breitkopfs letter presses,28 and in light of the parallel example of Bachs Art of Fugue,
25 J. S. Bach. Neue Ausgabe Smtlicher Werke. Serie IX/Band 1, ed. Wisso Weiss, Katalog der Wasserzeichen in Bachs Originalhandschriften. Textband (Kassel: Brenreiter, 1985), 14. A paper with even larger dimensions, Imperialpapier, whose sheets measure no less than 57 3 78 cm., was not commonly available. 26 a) unicorn rampant containing the letter L; b) cursive monogram EL. 27 The same paper is encountered in one of the proof copies, A 15, of Bachs Clavierbung III. See Gregory Butler, Bachs Clavier-Ubung III: The Making of a Print (Durham, N.C.: Duke University Press, 1990), 2829. 28 The evidence for this is furnished by an entry in one of Breitkopfs accountbooks dated 10 July 1747. See Bach-Dokumente, Supplement zu Johann Sebastian Bach. Neue Ausgabe smtlicher Werke. Band 2. Fremdschriftliche und gedruckte Dokumente zur Lebens-geschichte Johann Sebastian Bachs, ed. Werner Neumann and Hans-Joachim Schulze (hereafter Dok II) (Kassel: Brenreiter, 1969), 436.

313

t he j ou r na l o f mus ic o log y
whose title/dedication and musical text were each printed on the same paper, it is likely that at least those music plates printed on paper v were pulled in one of the small Leipzig copper presses that worked for Breitkopf. Paper w29 comes from the mill of David Hertel in Auerbach near Zwickau. Although this paper is documented in Leipzig during the years 172950, it was also widely used by printers in the area of Halle/ Saale. Paper x30 is from the mill of Johann Christian Keferstein in Croellwitz, a small town situated on the left bank of the Saale River on the northwest outskirts of Halle. I have not found this paper in any Breitkopf publication, nor have I been able to nd it in any other Leipzig publication from the period (although my survey has not been exhaustive) and I suspect that this small mills limited production went to satisfy the needs of local printers. Because of the close proximity of Halle to Leipzig, the possible use of this paper by Leipzig printers cannot be ruled out entirely. My detailed examination of paper z con rms Wolff s suspicion that it is not distinct but rather the same as paper w.31 Not only are papers v and w of different provenance, but signi cantly they are of contrasting quality as well: Paper v is of high quality, whereas paper w is a mediumquality Postpapier.32

314

The Printing
The distribution of the various papers in exemplars from the main printing run is given in Table 1.33 For all surviving exemplars of each of the ve printing units in the main printing run (with the single exception of printing unit B in exemplar A 10, which will be discussed separately below), the distribution of papers is identical. Notice that paper v occurs in every leaf of a given printing unit only in printing units A, B, and D, precisely those units that were also printed on the luxurious oversize paper y of the dedicatory exemplar. In both printing units E and C, at least one leaf is printed on a different paper, w in the case of the former and both x and w in the case of the latter. If this merely represented the mixing of various papers in the course of the main printing run, one would expect distinct and largely random distributions of the different papers in each of the surviving
29 a) crowned posthorn in coat of arms, suspended four mark with the letters AB; b) monogram DH. 30 a) crowned monogram ICK; b) blank. 31 Wolff, NBA 8/1 KB, 50, 52. For this reason I have subsumed paper z under paper w. 32 Ibid., 50. 33 This takes as its point of departure the table provided by Wolff, NBA 8/1 KB, 51.

butler
FIRST INSTALLMENT A. Title; Dedication (oblong) 1r. Musicalisches / Opfer / . . . 1v. __ 2r. Allergndigster Knig, . . . 2v. bewerkstelliget . . . SECOND INSTALLMENT

y,v

B. Three-Part Ricercar; Perpetual Canon (oblong) y,v _ __ 1 Ricercar. mm. 160 y,v _ y,v _ 2 3 4 cont. mm. 60110 cont. mm. 111156 cont. mm. 157185; Canon perpetuus super Thema Regium __ E. Six-Part Ricercar; Canons (oblong) w _ 3 cont. mm. 3351 4 cont. mm. 5167

E. Six-Part Ricercar; Canons (oblong) _ 1 Ricercar a 6. mm. 117 2 cont. mm. 1833 _ _ 5 6 7 8 cont. mm. 6779 cont. mm. 8095 cont. mm. 96103; Canon a 2.; Canon a 4.; J. G. Schbler sc. __

v v

315

y,v

D. Various Canons; Canonic Fugue (upright) Regis Iussu Cantio Et Reliqua Canonica Arte Resoluta 1 Canones diversi super Thema Regium; Fuga canonica . . . mm. 19 2 cont. mm. 9178 __

C. Trio Sonata; Perpetual Canon(upright) - Sonata / sopril Soggetto Reale - __ 1 Traversa; Largo; Allegro mm. 169 4 cont. mm. 47113; Canon perpetuus 3 2 1 4 3 2 1 4 Andante; Allegro mm. 147 cont. mm. 69172 Violino; Largo; Allegro mm. 157 cont. mm. 53113; Canon perpetuus Andante; Allegro mm. 153 cont. mm. 58172 Continuo; Largo; Allegro mm. 160 cont. mm. 47113; Canon perpetuus

3 Andante; Allegro mm. 146 2 cont. mm. 61172 - __ - __

t he j ou r na l o f mus ic o log y
exemplars (as in Bachs Clavierbung IV, for example).34 Here, as in the main printing run of Bachs Clavierbung III, however,35 the distribution of the three different papers is identical in all surviving exemplars of the printing units in question. The distribution of the papers strongly suggests that for the main printing run, the plates were pulled in two separate installments. Printing units A, B, D and leaves 1, 3, and 4 of printing unit E were pulled on paper v in one installment, and leaf 2 of printing unit E plus leaves 2 through 7 of printing unit C 36 were pulled on paper w in a second separate installment.37 A close examination of the text of printing unit E gives added support to the conclusion that the plates for the second leaf (pages 3 and 4) were pulled as part of a distinct installment of the main printing run. Of the twelve manuscript corrections38 to the seven pages of musical text in this printing unit, it cannot be merely a coincidence that all but one occurs between the second half of measure 51 and the rst half of measure 67, precisely that segment of the musical text printed on page 4, the verso face of leaf 2. At the same time, the plate for this page is the only one in this printing unit upon which corrections were not made directly by the engraver after proofreading. These two interrelated details indicate that for this page only, the engraving errors on the plate were never erased and the corrected readings reengraved. This suggests that Bach never saw a proof copy of this particular page or if he did, that there was no time to engrave the necessary corrections, which then had to be entered later by hand. The printing history of this plate is in stark contrast to that of the other six plates in printing unit E.
34 For a cursory discussion of the papers in Clavierbung IV, see J. S. Bach. Zweiter Teil der Klavierbung, Vierter Teil der Klavierbung, Vierzehn Kanons, ed. Walter Emery and Christoph Wolff, Neue Bach Ausgabe smtlicher Werke, Serie V/Band 2. Kritischer Bericht (Kassel: Brenreiter, 1976), 9293. 35 For a full discussion of the distribution of papers in Clavierbung III, see Butler, J. S. Bachs Clavier-Ubung III, 2729. 36 For a possible explanation as to why the title wrapper of printing unit C was printed on paper x, see below. 37 It could be argued that the distribution of papers in the various exemplars from the main run of the original print came about because the printing units were farmed out to different presses at the same time because of the pressure of the approaching publication date. This would, in turn, suggest that there was only a single installment of the main printing run. However, in such a scenario the pages from one of the printing units would surely never have been divided up so consistently in all the exemplars of any given printing unit as is the case for printing unit E. In fact there was pressure in the face of two quite different printing deadlines, rst for the printing of the dedicatory exemplar in July 1747 and second for the publication of the collection in time for the opening of the Michaelmas Fair at the end of September. 38 For a list of the manuscript corrections in the printed text of the six-part ricercar, see Wolff, NBA 8/1 KB, 56.

316

butler
There are two plausible explanations for this discrepancy. Bach, on seeing the proof copy of the six-part ricercar, decided to revise a passage on page 4. This revision may have been extensive enough that the erasure of the original reading and the reengraving of the revised reading was judged to have been unfeasible. Another possibility is that the plate for page 4 was damaged either during the process of erasure, (re)engraving, or printing, a not infrequent occurrence. In either case, the plate in question would have to have been discarded and a new plate engraved to take its place. As a consequence it would not have been possible to pull this plate along with the others in this printing unit as part of the rst installment of the main printing run, but only sometime later in a second installment. Leipzig was arguably the principal book printing and publishing center in Germany at the time, and its numerous printing presses are certain to have been swamped with work, particularly during the periods leading up to the major trade fairs held there. I suggest that the necessity of having to have page 4 of printing unit E reengraved delayed the completion of the project as originally planned. By the time the reengraving had been carried out, Bach may well have been unable to nd a copper press available in Leipzig for the pulling of the second leaf of printing unit E and was forced to turn to another press,39 on the evidence to be drawn from paper studies (see below), possibly in nearby Halle,40 for the pulling of the second installment of the main printing run. Its inclusion in the rst installment of the main printing run would suggest that printing unit E had been engraved by the time the dedicatory copy was sent off to Potsdam. When it became apparent that printing unit E could not be included in the dedicatory copy as originally planned, Bach may have decided to substitute printing unit D. Printing unit D then would have been engraved later, sometime after the completion of printing units B and E, possibly during the hiatus in production,
39 Bach changed not only printing shops but engraving shops as well for the second installment of the printing run of Clavierbung III. See Butler, J. S. Bachs Clavier-Ubung III, 21ff. According to Wolff, the paper proves that the printing was done in one working procedure and, surprisingly, not at two different places but obviously at one. This glosses over the important evidence of the break in paper and the equally important evidence to be drawn from it, as does his statement that the paper for A (set up in type by Breitkopf in Leipzig) and for the musical sections (engraved by Schbler in Zella) is identical in all copies, including the dedication copy. See Wolff, New Research, 252. In fact Wolff s reference to musical sections applies only to printing units B and D. 40 It may be signi cant that it was at about this time that Wilhelm Friedemann Bachs Sonata in E (preface dated 8 January 1748) which, like his fathers Musical Offering, had been engraved by the Schbler brothers in Zella, was being printed in Halle. Having their plates pulled in the same press would have been advantageous nancially for both father and son. An added incentive for the possible change of locale may have been the reduced cost of printing in the smaller town.

317

t he j ou r na l o f mus ic o log y
by which time Johann Georg Schbler appears no longer to have been available for the engraving. If so, then it would seem that it was the unforeseen necessity of having to have page 4 of printing unit E reengraved that extended the life of the project and that allowed Bach to expand his original, more modest conception of the collection. If my conclusion that one plate of printing unit E was unavailable for printing at the time the rst installment of the main printing run passed through the press is correct, then neither it nor printing unit C (which is printed on the same paper) can have been pulled along with printing units A, B, and D and printed on the luxurious oversized paper of the dedicatory copy unless we are to believe that the dedicatory copy was not pulled until the second installment of the main printing run went through the press in September 1747.41 Bachs expressed wish to put down on paper and afterwards to have engraved42 the fugue which he had improvised before the king at the fortepiano during his visit and sent off to Potsdam was a matter of public record, and Bach would have been under considerable pressure to make good on his promise. I suggest that only those printing units which were ready for printing in their entirety in July, 1747,43 (A, B, and D) were ever pulled on the luxurious, oversize paper y. If my conclusion is valid, then Spitta was correct on this matter: Only his sections 1 and 2 were originally sent to the king along with the title page and dedication. If one accepts the compelling evidence pointing to a main printing run divided into two distinct installments and the conclusion drawn from itthat only printing units A, B, and D were originally sent to the kingthen the possibility cannot be ruled out that Bach sent the remaining two printing units, C and E, to complete the dedicatory copy later around the time that the collection as a whole was published.44
41 We know from the publication announcement that the last plates to be pulled had passed through the press by 30 September 1747. See Bach-Dokumente, Supplement zu Johann Sebastian Bach. Neue Ausgabe smtlicher Werke. Band 3. Dokumente zum Nachwirken Johann Sebastian Bachs 17501800, ed. Hans-Joachim Schulze (hereafter, Dok III ) (Kassel: Brenreiter, 1972), Nachtrge zu Band II, 656. 42 zu Papiere bringen, und hernach in Kupfer stechen lassen See Dok II, 43435. This wish, speci cally mentioned in the newspaper report of Bachs performances before the king, had become a promise in the announcement of the works publication with its reference at the outset to das unterm 11. May a.c. in denen Leipziger, Berliner, Franckfurter und andern Gazetten versprochene Koenigl. Preussische Fugen-Theme. See Dok III, 656 43 The dedication is dated 7 July 1747. 44 The appearance of a third engraver along with the other two for the engraving of a single printing unit, C, offers a sharp contrast to the modus operandi adopted for the engraving of the other printing units, which are each the work of a single engraver. The deployment of a three-man team of engravers suggests that the plates for printing unit C had to be engraved in some haste, probably because of the deadline for the publication of the collection in time for the opening of the Michaelmas Fair.

318

butler
That he sent ordinary exemplars would suggest either that paper y was no longer available (as Spitta concluded) for the second installment of the main printing run, or that the bed of the copper press in which the second installment was pulled was not large enough to accommodate these oversize leaves.

The Wrapper Theory


Because it is bound up with paper formats and printing conventions, Wolff s wrapper theory must now be addressed. This theory holds that the three uncut folio leaves (printing unit A, the title page to printing unit C, and printing unit D) were originally intended to serve as wrappers for printing units B, the musical text of printing unit C, and printing unit E respectively, thus forming three distinct fascicles. Wolff s belief that paper y was available only as single sheets led him to conclude that the wrapper theory applied only to the normal exemplars and not to the dedicatory copy. This, in turn, suggested a chronology of events in which the printing run for the ordinary exemplars preceded and was independent of that for the dedicatory exemplar:
The ne paper used for this special copy was available only in single sheets and not in bifolios. Thus [in the case of the dedicatory copy] there was no possibility for forming three fascicles with separate title covers as in the ordinary copies. We see now the reasons for the different makeup and particularly for the handwritten inscriptions that are to be found only in the dedication copy. Since the fascicle structure was inapplicable and consequently the acrostic could no longer function as subtitle for the third fascicle, Bach put the acrostic on the blank folio 1 recto of section B, at the beginning of the entire work. He obviously did not want to abandon the brilliant idea of the acrostic. Because the paper strip with the engraved acrostic was disproportionately small for the large folio, the calligraphic handwriting was substituted. As a result of this shift, folio 1 recto of section D received a new inscription avoiding any reference to the six-part ricercar, namely, Thematis Regii Elaborationes Canonicae.45

319

This reconstruction can now be seen to be untenable. The accumulated evidence suggests that the dedicatory copy comprising printing units A, B, and D was pulled and dispatched to Potsdam some two months before the expanded version of the collection became available following the second installment of the main printing run. Because printing unit E had not yet been printed in its entirety when the initial installment of the dedicatory copy was sent to the king, there can be no
45

Wolff, New Research, 251.

t he j ou r na l o f mus ic o log y
question of the acrostics being shifted from the blank recto face of the rst leaf of printing unit D to the analogous page of printing unit B at this time. The changes Wolff outlines took place, but in reverse order: only in the nal stages of the production of the original print just prior to its publication did it occur to Bach to print the acrostic on labels and switch its position to serve as the general title for printing unit D. Further, since it has been shown that paper y was available as bifolios, then the wrapper theory would also have been applicable to the dedicatory copy, at least insofar as the rst fascicle was concerned. However, the argument that Bach intended printing unit A to function as the wrapper for printing unit B in the dedicatory copy is dif cult to uphold. Printing unit B in the dedicatory copy carries its own general title in the form of the acrostic and (like the inscription on printing unit D of the dedicatory copy) was surely conceived originally to serve as a general title referring to the speci c contents of printing unit B: Regis Iussu Cantio [three-part ricercar] Et Reliqua Canonica Arte [perpetual canon] Resoluta. The Iussu of the acrostic refers directly back to the parallel Befehl of the dedication, and the object of the royal command was the three-part, not the six-part ricercar which at any rate was not included in the rst installment of the dedicatory copy. Given the acrostic that clearly indicates the contents of the dedicatory copy, a wrapper would have been super uous. Surely printing unit A must be seen as the title page to the collection as a whole and not just one of its components. Following the same reasoning, it is hard to imagine how printing unit D could ever have been intended to function as the wrapper to printing unit E. The inscription on the blank recto face of leaf 1 of printing unit D in the dedicatory copy, Thematis Regii Elaborationes Canonicae, clearly refers only to the contents of that printing unit, the Canones diversi super Thema Regium. There is no way of construing it to refer to the contents of printing unit E as well. Further, while the wrapper theory is attractively simple and elegant, in attempting to be all-encompassing it fails to take into account the distinct natures of each of the three wrappers in question. The wrapper enclosing the parts of the trio sonata differs fundamentally from the other two in that it forms an integral part of its printing unit rather than a separate printing unit. It is the only one of the three wrappers devoid of printed text beyond the title, and it was clearly intended by Bach to act as a wrapper from the outset. Here its function as wrapper is entirely in keeping with musical tradition and Bachs own normal practice and is not in question. The bifolio carrying the title of the collection and the dedication is printed by means of movable type and must therefore be considered in light of the tradition of the printed book. Like all leaves of printed

320

butler
books published at that time, this large oblong folio leaf would as a matter of course have been cut along the fold by its purchaser so that when bound (along the short side) the rst of its two leaves could be opened and the inner pages (leaves 1 verso and 2 recto), on which the dedication is printed, read. Of all of the surviving exemplars of this printing unit, only one, A 9, remains in its original uncut state. I would submit that in most if not in all of the other exemplars this leaf was cut by the original owner routinely not long after purchasing the collection and having it bound. 46 The cut exemplars of this printing unit thus are representative of the normal state of affairs, whereas the single uncut exemplar is the exception. Cutting this folio leaf did not have the effect of destroying any original layout, nor does it constitute an act of mutilation. It simply permits the leaves to be opened and their contents read once the book has been bound. Further, there is no pressing necessity for a bifolio to function as a folder to keep single leaves together. All the original editions of the works of Bach in oblong format were printed on single leaves, collated and sold unbound. The musical text of Bachs Art of Fugue was similarly printed on single leaves in oblong format, and the work also includes a separate printing unit whose format and disposition is identical to that of printing unit A, consisting of a title page and preface printed by means of movable type on a bifolio in oblong format. Yet there has never been any question that this served as a wrapper to contain the single leaves of engraved music. Surely the bifolio carrying the title and dedication of the Musical Offering, like the parallel printing unit in Bachs Art of Fugue, would have appeared on the book stands in October 1747 as a separate uncut printing unit, not functioning as a wrapper for printing unit B. In conclusion, I would submit that the only wrapper involved here was that which held the somewhat disparate contents of the dedicatory exemplar as a whole. Two empty leavesone preceding the title page and the other following the empty verso face of the fth oblong sheet in Spittas section 1act as endpapers today of the dedicatory copy. They are leaves of the same paper (y) on which the three printing units of the dedicatory copy are pulled, and indeed Wolff suggests that originally they may have functioned as some kind of endpapers for the printed pages.47 Originally these two single leaves probably formed a single bifolio, and I suggest that this blank bifolio probably acted as a folder to keep together and protect the other printing units of the dedicatory exemplar.
46 According to Wolff, all other copies of section A [printing unit A] were cut on the fold for binding purposes in the later eighteenth, nineteenth, or even twentieth century, thus destroying the original layout. Ibid., 249. 47 Wolff, NBA 8/1 KB, 5859.

321

t he j ou r na l o f mus ic o log y
The Printing
Before embarking on a discussion of the main printing run of the collection, I would like to focus brie y on the paper format employed. Because the entry of 10 July 1747 in the accounts of the Breitkopf rm includes the clear speci cation 200 Ex. Royl.48 it has quite naturally been assumed that the work was printed on Royalpapier. However as already noted above, paper y, the special oversize luxurious paper upon which the dedicatory exemplar is printed, is a Royalpapier. The paper employed for the main printing run is conspicuously smaller in its dimensions. In the only surviving uncut and unbound exemplar, A 9, the pages measure 23.5 3 37.0 cm.49 The uncut folio sheets would then have measured 37 3 47 cm., which corresponds almost precisely to the dimensions of the normal medium-sized sheets of paper manufactured in Bachs day, the so-called Median format whose dimensions are 38 3 48 cm.50 Was Breitkopfs accountant here perhaps referring to the paper upon which the dedicatory copy had been pulled, or might Bach perhaps have initially contemplated having all two hundred exemplars of printing unit A printed on Royalpapier and then changed his order having seen that the dimensions of Schblers plates were compatible with those of half sheets of paper in Median format or simply because paper y was too costly? The answer to this question must remain a matter for conjecture. Certainly having the main printing run pulled on paper of smaller format and inferior quality would have spared Bach considerable expense as publisher. What precisely is the relationship of the dedicatory exemplar to the ordinary exemplars of the rst installment of the main printing run insofar as the printing history of the collection is concerned? Following his normal procedure, once he had received the plates for printing units B and D, Bach would have had a proof copy of the rst installment of the main printing run pulled by the printer.51 Although this proof copy is no longer extant, evidence of corrections made directly on the plates52 indicates that Bach saw proofs and sent a list of corrections back to the engravers along with the plates. Once the plates had been corrected it would have made no sense for Bach to have had the dedicatory exemplar pulled in a separate printing run. This would have
Dok II, 436. Wolff, NBA 8/1 KB, 65. 50 Weiss, Katalog, 14. Wolff gives similar dimensions (ca. 40 cm. 3 50 cm. pro Bg.) as those for Royalpapier. NBA 8/1 KB, 47. 51 Such proof copies survive. For example, that for Clavierbung III is the source A 7. For a discussion of the role of this exemplar in the correction process, see Butler, J. S. Bachs Clavier-Ubung III, 6570. 52 For a discussion of these corrections, see Wolff, NBA 8/1 KB, 5355.
48 49

322

butler
been both time-consuming and costly. After the proof copy had been pulled and corrections made to the plates, each plate would have been pulled rst on the special oversized luxurious paper set aside for the dedicatory exemplar, y; immediately after, a predetermined number53 of copies were pulled on the normal paper chosen for the rst installment of the main printing run, v. Upon receipt of the printed pages from the rst installment of the main printing run, Bach must have assembled a clean, uncorrected correction exemplar54 to act as a control for the correction by hand, page by page, of the printed pages of the rst installment. The most likely candidate for this correction exemplar is the source A 15, the only uncorrected exemplar. After the necessary corrections were made by hand in the dedicatory exemplar, it was sent off to Potsdam. The ordinary exemplars of the rst installment of the main printing run may have been corrected at this time or later. It appears that Bach had not decided to include the six-part ricercar in the collection until after the penning of the dedication on 7 July 1747.55 However, it must have been written and engraved before the pulling of the rst installment of the main printing run, because printing unit E is printed (with the exception of its second leaf ) on the same paper as that used for the balance of this installment. The fact that the only exemplar in which it remains uncorrected is also that in which printing unit B remains uncorrected, A 15, strongly suggests a continuity in the printing and correction processes for these two printing units. The exemplars of printing unit E were duly pulled (again with the exception of the second leaf) as part of the rst installment. As demonstrated above, a problem with the plate for page 4 necessitated a second installment in the main printing run. Although Bach almost certainly intended originally to include printing unit E in the dedicatory exemplar, he was unable to do so because of the problem with the plate for page 4. The plates for printing unit E are the most painstakingly and thoroughly corrected by the engraver of any in the collection, probably because this printing unit was not produced under the same extreme time pressure as printing units B and D, and later, printing unit C. In stark contrast, page 4 gives every appearance of being the only page never to have been proofread by Bach, as stated above. Not only are the telltale
53 For printing unit B we know that this number was 100. See Bachs letter of 6 October 1748 to Johann Elias Bach, Dok I, 11719. 54 There are two such uncorrected exemplars for Clavierbung III, one of which clearly served the function of correction exemplar. Butler, J. S. Bachs Clavier-Ubung III, 7071. 55 The obvious septocentricity of the date of the dedication, 7 July 1747, is arti cially contrived. It may be that the dedication was written and printed along with printing units B, E with its missing leaf, and D before that date, with a view to the delivery of the dedicatory exemplar then or soon after.

323

t he j ou r na l o f mus ic o log y
signs of engraved corrections absent here, but with its ten manuscript corrections it quali es as the most heavily corrected page in the collection. The reengraving of this plate would have taken place later before the printing of the second installment of the main printing run, perhaps at the same time as the engraving of the parts for the trio sonata. It is a little puzzling that a different paper, x, was used for the printing of the title wrapper to printing unit C than that used for the music plates of the second installment, paper w. This suggests that the title wrapper was printed at a different time, perhaps even in a different press, than the parts to the trio sonata and the second leaf of printing unit E. There remains the discussion of the printing of the acrostic and the title of the sonata. Wolff s explanation for why the acrostic was not printed directly on the blank outer face of printing unit Dbut rather on a paper strip which was then glued onto itis entirely convincing. He suggests that it was engraved on a smaller plate,56 but as he notes correctly, nowhere are plate borders in evidence57 on this paper strip as they are in the case of the engraved title to the sonata. This would suggest that the strips were cut out with scissors either along or inside the outline made by the plate edges. This makes for an excessively narrow plate if we use Bachs printed canon BWV 107658 as a basis for comparison. I would like to suggest that there is no sign of plate borders for the simple reason that there never were any. The acrostic was not printed from an engraved copper plate as were the pages of musical text but, like the so-called Hudemann canon BWV 1074,59 from a woodblock. Because in this process the letters were raised rather than excised, the block borders leave no outline. This explanation is entirely in keeping with the fact that the Schblers were trained not only in iron cutting but also in carving gunstocks. It also represents an elegant solution to the costly and time-consuming expedient of pulling a narrow and hence fragile plate over and over. The woodblock could be re-inked quickly and the acrostic printed by hand by Bach himself or someone from his circle many times on the same large sheet of paper that was subsequently cut into strips. Bach as publisher thus not only was spared needless time and expense but also would have had direct control over this delicate but important step in the production of the print.
See Wolff, New Research, 250n31. Plattenrnder sind nirgendwo erkennbar. Wolff, NBA 8/1, KB, 47. 58 The dimensions of this diminutive plate are 8.3 3 10.2 cm. Ibid., 22. 59 In Johann Matthesons Der vollkommene Capellmeister, in which this canon appears, no plate edges are visible in any exemplar. For a discussion of the engraving and printing of this edition of the canon, see Gregory G. Butler, J. S. Bachs Hudemann Canon BWV 1074: Notes on the Engraving and Printing History, Bach 24 (1995): 110.
56 57

324

butler
The Marketing
That Bach originally had 200 exemplars of printing unit A run off, whereas the main printing run of the four printing units of musical text amounted to 100 exemplars, has been explained by scholars as a shrewd calculation on Bachs part.60 The evidence that appears to establish the second number is provided by the postscript to a letter written by Bach on 6 October 1748 to his cousin and former amanuensis, Johann Elias Bach:
I cannot oblige you at present with the desired copy of the Prussian Fugue (Preussische Fuge), the edition having been exhausted just today, since I had only 100 printed, most of which were distributed gratis to good friends. But between now and the New Years Fair I shall have some more printed, and if then my honored Cousin is still of a mind to have a copy, he need only give me notice on occasion, sending me a Thaler 61 at the same time, and his wish shall be ful lled.62

David, in referring to this document, allows that the title Preussische Fuge may indicate that Bach did not intend to have the Sonata and Canons reprinted with the fugue or fugues.63 In fact, there is no reason to believe that Johann Elias was requesting any composition from the collection other than the three-part ricercar, or that Bach was referring to anything other than printing unit B containing the three-part ricercar when he says that the edition was exhausted just today. Nor is there any reason to doubt Bachs word that he would have a few more copies [i.e., of printing unit B] printed. In fact, evidence of a reedition of printing unit B does exist.
60 It is explained by Bachs having anticipated a reprinting of the collection. Whereas the copper plates for the musical text could be retained by the composer or the printer and could be pulled over and over, the text of printing unit A, which was set in movable type, would have been broken up soon after the printing; its two pages would have to have been reset by the compositor for any subsequent reprinting (as in the rst and second editions of Bachs Art of Fugue) at great cost to the composer as publisher. 61 One Thaler, the advertised price of all ve printing units of the collection, might seem to be an excessive sum for Bach to be charging a former amanuensis and family member for a single printing unit, particularly because he mentions having distributed gratis to good friends exemplars from the original printing run. However, one should remember that whereas the original edition may have been subsidized in part and Bach may have been left with remainders that he was unable to sell (as was clearly the case with other of the original prints of his works), this reprinting, a costly and risky venture, would almost certainly have been at his own expense. One should also not forget Bachs propensity for frugality, even tight- stedness, in nancial matters as a factor here. 62 Dok I, 11719. The English translation given here appears in The New Bach Reader, ed. Hans T. David and Arthur Mendel, revised and expanded by Christoph Wolff (New York: W. W. Norton, 1998), 234. 63 David, J. S. Bachs Musical Offering, 10.

325

t he j ou r na l o f mus ic o log y
For all surviving exemplars of any given printing unit, the papers and the distribution of those papers are the same, with a single exception. Exemplar A 1064 includes printing units A, B, and E. The papers for printing units A and E and their distribution in printing unit E in these exemplars are the same as for all other exemplars of those printing units. However, whereas all other 12 exemplars of printing unit B from the main printing run are printed on paper v, printing unit B in this exemplar is on paper w. Also, with the exception of A 15 (whose special identity as a correction exemplar has been detailed above), this is the only exemplar from the main printing run in which none of the six essential manuscript corrections in printing unit B were made. The accumulated evidence would suggest that before the end of 1748, in advance of the New Years Fair of 1749, Bach had a few additional copies of printing unit B (for which there seems to have been a particularly heavy demand) run off in the same copper press he had engaged for the second installment of the main printing run. He could then have offered them for sale along with the other printing units, which would still have been in stock. If Bach was referring only to printing unit B in the letter quoted above, did he also have fewer than 100 exemplars of other printing units run off? That the rst installment of the main printing run pulled in the copper press (printing units B, E with its missing leaf, and D) was of 100 exemplars explains why the same number of exemplars of printing unit E (14) survive as do exemplars of printing unit B.65 There is nothing to substantiate the theory that fewer copies of printing unit C survive (3) than for the other printing units because there was more wear and tear on it as a piece of chamber music. Based on the numbers of Bachs other original editions that survive, I think one can realistically reckon on a survival rate of about 10 to 12 percent for each of the music printing units. If this is so, then only about 30 exemplars of printing unit C would have been run off in the second installment of the main printing run. Joel Sheveloff 66 has suggested that the various printing units of the collection were published individually and sold separately. The publication-by-installment theory is not supported by paper studies that
64 This is one of the three exemplars of the original print in the music collection of the Oesterreichische Nationalbibliothek, A:Wn S. H. J.S. Bach 101. In this exemplar, the acrostic (which in all other normal exemplars is printed on a strip of paper and glued onto the empty recto of the rst leaf of printing unit D) is entered by hand on the empty recto of the rst leaf of printing unit B, as in the dedicatory exemplar. This would suggest that the woodblock from which it was printed was no longer available. 65 For an explanation of why only about half this number of exemplars of printing unit D survive, see below. 66 Joel Sheveloff, Quaerendo Invenietis (M.A. thesis, Brandeis Univ., 1964), 11ff.

326

butler
demonstrate clearly that the music printing units were printed in two distinct installments, not four, and that even these two installments do not divide neatly between two printing units but rather overlap. In responding to Sheveloffs theory Wolff quite reasonably poses the question: But why then is the composers name found only on the title page and not at the beginning of each section? The distribution and sale of pieces without mention of the author is unthinkable.67 I believe that Bach was an even more astute and creative businessman than we give him credit for. I think that he calculated correctly that the two ricercars, simply for practical and nancial reasons, would be eminently more saleable than the canons or the sonata. For this reason he initially had only half the number of exemplars of printing unit D run off as of printing units B and E. When printing unit C was nally produced, Bach decided to have even fewer exemplars run off than for printing unit D, probably because at 12 pages plus a title page it was twice as costly to produce as the next largest printing unit, printing unit E. He was allowing for the eventuality that any given buyer might choose to purchase any of these four printing units singly, any two of the four, any three of the four, or all four. He simultaneously hit upon the idea of including in any purchasewhether of one, two, three, or four printing unitsan exemplar of the title page and dedication. With this in mind he had twice as many exemplars of printing unit A run off, not because he had already allowed for a future second printing of another 100 exemplars of all of the four music printing units (a very costly venture), but because he had decided to include a copy of printing unit A with each purchase. This scenario solves Wolff s problem with the installment theory, for any printing unit(s) sold would have been accompanied by the title page with the composers name displayed prominently. It would also go a long way toward explaining why each printing unit has its own pagination. Following this line of reasoning, each printing unit would then have carried its own individual price. Of the surviving exemplars, only two are complete and both represent extraordinary, arti cial cases. The rst is the special exemplar for the dedicatee that constituted a complete copy, although its printing units arrived in Potsdam in two installments each with its own paper format and quality. The second, lacking only printing unit A, is A 13, a complete exemplar assembled by the Hofkapellmeister Giovanni Battista Pauli for Padre Martini in Bologna in 1750.68 Each of the other exemplars is incomplete, not because printing units have been lost but because in
Wolff, New Research, 248. For his interpretation of the disposition of this exemplar, see Marissen, More Source-Critical Research, 1314.
67 68

327

t he j ou r na l o f mus ic o log y
virtually every case, I would submit, they represent the particular printing unit(s) chosen by the original purchaser. Taken together, they re ect the printing units that most appealed to purchasers.

Conclusion
In the foregoing study it has not been my intent either to present a new ordering scheme or to deny that Bach ever had a coherent disposition in mind for the collection as a whole. I have simply suggested, based on my interpretation of the available source evidence, the order in which the various printing units were produced: B/E, D, C. However, I believe that the sequence of events I have outlined does shed some light on the question of ordering. For each of the four music-printing units the Abklatschvorlage 69 engraving technique was employed in preparing the plates.70 The fact that in this procedure the engravers traced directly onto the plates, page by page, a fair autograph supplied by the composer means that within any given printing unit of this collection, the ordering necessarily re ects that intended by Bach at the time he submitted his engravers copy. Therefore, any hypotheses as to the intended ordering must be limited to the ordering of the printing units, and not to that of the compositions each contains. I interpret the signature J. G. Schbler that appears at the foot of page 7 (the concluding page of printing unit E) as a gesture of nality on the part of the engraver, a concluding ourish. It suggests to me that Johann Georg Schbler engraved printing units B and E in that order and that the completion of work on these two printing units marked the end of the project not only as originally projected in the engravers mind but in that of the composer as well. In Bachs initial conception
69 In this technique, sheets with music on one face only were soaked in oil to render them transparent. These transparencies were then applied with the music side down to varnished copper plates upon which the music was then traced in mirror image. The plates were then incised by the usual process of etching. This is the technique adopted by the Schbler brothers for the three engraving projects they undertook for J. S. Bach, and indeed it is the technique in evidence in all of their surviving music engraving. 70 Marissen has suggested that some plates may have been engraved by the process of freehand engraving. More Source-Critical Research, 1625. The consistency and homogeneity in the general appearance of the corpus of the Schblers engraving argues against his interpretation. It should be remembered that the Schbler brothers were trained not as engravers but rather in the craft of gun stock carving and ornamental iron incising. (See Wiemer, Johann Georg Schbler, 8587, 9095.) Lacking the necessary expertise required for mechanical or freehand engraving of music, they naturally applied the same technique for engraving music on copper plates as they used for incising ornamental patterns on trivets, gun barrels and the like. Virtually all of the details of engraving Marissen cites in support of his argument for the use of a method of engraving other than the Abklatschvorlage method can be interpreted in other ways.

328

butler
then, the collection was to include only printing units A, B, and E: the title/dedication and the three- and six-part ricercars with their appended canons. The two ricercars correspond to the two pieces Bach elaborated from the theme provided by the kinga fugue improvised by Bach on the fortepiano the rst night, and another not on the kings theme in six parts that he improvised laterthe two works that he expressed a wish to write down and subsequently have engraved and sent to the king. Printed in oblong format, like the title page and dedication, aesthetically these three printing units would have represented a harmonious ensemble. I would maintain that these three printing units in the order A, B, and E represent the early version of the collection. Whether the diverse canons of printing unit D were composed while the two ricercars were being engraved and printed or whether they were composed only laterwhen it became apparent that printing unit E could not be included in the dedicatory copy as originally planned is not clear. As compensation, Bach may have decided to substitute printing unit D for printing unit E. Bach had already adopted the concept of the canonic appendix in the Fourteen Canons at the end of the Handexemplar of the Goldberg Variations and in the last variation in the printed version of his Variations on Vom Himmel hoch, and he would resort to it once again in the canons that conclude the early (autograph) version of the Art of Fugue. I would argue that he saw the diverse canons of printing unit D in the same light; that is, as a pendant to the collection as a whole, a re ection on the macro level of the canons that appear as appendices in each of the other three music printing units. This leaves for consideration only the trio sonata and appended canon of printing unit C. This was the last of the printing units to be produced, and its contents were also likely the latest works in the collection to be composed. However, Bach can surely never have intended that it serve as the conclusion to the collection, for as I have already speculated, this position would have been reserved for the canonic appendix, printing unit D. It is equally unlikely that Bach ever considered that it might open the collection. Like the diverse canons of printing unit D, the trio sonata with its appended canon seems to have been an afterthought, though one highly a propos of the dedicatee of the Musical Offering, the utist king, and of the regular chamber music concerts in which he performed with his court musicians. Certainly its upright format is the same as that of the diverse canons of printing unit D, and the pieces in each of these printing units include instruments in addition to keyboard, so that one might make the argument that they belong together, just as do printing units B and E. The consequence of this line of reasoning, the placement of printing unit C between printing units B and E and printing unit D, yields

329

t he j ou r na l o f mus ic o log y
the same disposition given in the newspaper advertisement announcing the publication of the collection, which states that the elaboration consists 1.) of two Fugues, one with three, the other with six obbligato parts [printing units B, E]; 2.) of a Sonata for Transverse Flute, Violin, and Continuo [printing unit C]; 3.) in various Canons, including a Fuga canonica [printing unit D].71 Marissen makes a strong argument for this ordering of the printing units,72 and in support of his argument it is signi cant that the newspaper advertisement clearly considers printing units B and E to be a unit, as the evidence I have presented also suggests. This also supports my view that they, along with printing unit A, constitute Bachs early concept of the work. Whoever provided the disposition for the advertisementwhether Bach himself, as is likely,73 or a member of his circlede nitely viewed B and E as a single unit, and this should make us wary of any ordering scheme that breaks them up. My experience with the other original prints from the composers last decade (more speci cally, his Canonic Variations and Art of Fugue) suggests that just as Bach was constantly revising musical readings and embarking upon further expansion of collections both manuscript and printed, so too was he rethinking the ordering schemes of those same collections. Having encountered a problem with the engraving of the Canonic Variations, during the ensuing hiatus in the production of the original print Bach expanded the collection by composing the augmentation canon. This ultimately forced him to rethink the original ordering scheme of the collection,74 much the same way as his addition of two canons to the original two in the initial conception of the Art of Fugue changed the ordering scheme of that collection. It is entirely possible, therefore, that once he had made the decision to add the diverse canons of printing unit D, and then the trio sonata and perpetual canon of printing unit C to his original conception of the Musical Offering, he similarly changed his mind about the ordering of the collection as a whole. Bachs decision to label printing unit D for the exemplars of the main printing run with the acrostic originally inscribed on printing unit B in the dedication copy may be an indicator of such a change of mind.

330

71 Dok III, 656: Die Elaboration bestehet 1.) in zweyen Fugen, eine mit 3. die andere mit 6. obligaten Stimmen; 2.) in einer Sonata, a Traversa, Violino e Continuo; 3.) in verschiedenen Canonibus, wobey eine Fuga canonica be ndlich. 72 See Marissen, More Source-Critical Research, 1216. 73 Marissen sees no reason to assume that the wording is not Bachs. More SourceCritical Research, 13. 74 For a detailed discussion, see Butler, J. S. Bachs Clavier-Ubung III, 10809. See also Gregory Butler, J. S. Bachs Kanonische Vernderungen ber Vom Himmel hoch (BWV 769). Ein Schlustrich unter die Debatte um die Frage der Fassung letzter Hand BachJahrbuch 84 (2000): 934.

butler
I agree with Wolff that for purely practical purposes of marketing the product and performing it, not only is there no apparent ordering scheme for the collection, but also that its piecemeal distribution and dispersal seem to have been almost entirely haphazard and a function of choice in a free-market economy. That Bach had in mind a cogent abstract disposition scheme for the collection as a whole cannot, of course, be ruled out. My principal aim here has been neither to verify nor to disprove any of the various ordering schemes proposed by Bach scholars but rather to clarify certain aspects of the production of the collection, most importantly its complex engraving, printing, publishing, and marketing history. University of British Columbia

ABSTRACT
Evidence from the original printed edition not only supports Philipp Spittas theory that Bach sent the print of his Musical Offering to Frederick the Great in Potsdam in two separate installments but further suggests that the work was also printed in two distinct sections, the rst in Leipzig and the second possibly in the region of Halle. In Bachs earliest concept of the work, it was to have included only the three- and six-part ricercars; he subsequently enlarged the scope of the collection. A detailed examination of the engraving suggests why this was so, and it also reveals that a third Schbler brother was involved in the engraving of the plates. Finally, the heterogeneous makeup of the surviving exemplars of the print hints at a novel strategy adopted by Bach for the marketing of the work. 331

You might also like