3..
an

:3...
3 ’1! .

 

 

 

 

€ fiém.
....Iar\.b..u...:v
p flu. 5“:

N. xii-.5
.1: S

”3.x...

D} T

. v
.vtflb......:
x

2n ‘.
.{Wxxki}
. I: .1 I...
In; .1).
.7. i..-

A. titan
xi? :3

I. ,
7.6.

s. e .I.§Ah1:
.311'! ‘ 3.35 :1
‘33.!!!» ~
#1.‘ 9!: ‘
'3: . :a t
I?! I: \ .

ntasiff»

.21 3.
95.1.4.3»!

1
i... .I‘z
9.. :5 2).. .9...
V :l’. . .
. ‘3: 5 r1;lx|i
;§.§5.qu 3)
... -\.I

)3»; ‘M
n 3. v .vadfl A
. . I: ,

,O’Wt-

W Illlllllllllllllllll

This is to certify that the

dissertation entitled

FROM CONTINUO TO OBBLIGATO CEMBALO:
A STUDY OF THE CHANGING ROLE OF THE HARPSICHORD
IN J.S.BACH'S HARPSICHORD CONCERTOS
AND SOLO SONATAS WITH OBBLIGATO CEMBALO

presented by

Robert James Galloway

has been accepted towards fulfillment
of the requirements for

 

Ph . D . degree in MuailegL

 

Major professor

Date 1-18‘96

MSU is an Affirmative Action/Equal Opportunity Institution 0-1277!

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to romovo this chockout from your record.
TO AVOID FINES rotum on or botoro dot. duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSUIoAn‘” “ ‘ ‘ 'j ' .—.—

FROM CONTINUO TO OBBLIGATO CEMBALO:
A STUDY OF THE CHANGING ROLE OF THE HARPSICHORD
IN J. S. BACH’S HARPSICHORD CONCERTOS
AND SOLO SONATAS WITH OBBLIGATO CEMBALO
By

Robert James Galloway

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

School of Music

1 995

 

ABSTRACT

FROM CONTINUO T0 OBBLIGATO CEMBALO:
A STUDY OF THE CHANGING ROLE OF THE HARPSICHORD
IN J. s. BACH’S HARPSICHORD CONCERTOS
AND SOLO SONATAS WITH OBBLIGATO CEMBALO
By

Robert James Galloway

One of the most significant and far—reaching musical developments of the
eighteenth century was reflected in the changing attitude toward the basso continuo or
thorough-bass. Retained as a prerequisite for the performance of all music since the
beginning of the seventeenth century, the continuo represented much more than a mere
performance practice, for in effect it controlled the very style and texture of virtually all
musical composition. So pervasive was this practice that the continuo role became the
principal function of the harpsichord in the period, and many tutors were written on the
subject of thorough-bass accompaniment. Despite such wide-spread acceptance,
however, adherence to basso continuo began to weaken in the mid-eighteenth century,
becoming nearly obsolete by the 1770’s and 1780’s in the face of a growing recognition
that its function had become superfluous.

The shift away from continuo practice was most apparent in chamber music with
smaller instrumentation. In the early stages of this transition appeared a number of
chamber works by J. S. Bach which not only illustrate the emerging obbligato cembalo
but also provide a means of observing the very process of transformation, since the

harpsichord part is derived to a considerable degree from what previously had been a

 

continue part. Parallel versions of both concertos and sonatas permit comparison and
shed light on the important process of adaptation. While the concertos generally are
more brilliant in execution than the sonatas, there are nonetheless common elements,
particularly with regards to the derivation and formation of the cembalo bass part.
Bach’s harpsichord concertos, BWV 1052-58, form the central core of this study
and reflect not only the emerging solo cembalo but also the early evolution of the
keyboard concerto. Ultimately this development, transmitted further via Bach’s sons
Philipp Emanuel and Johann Christian, was to be of far—reaching import to the shaping

of a new genre which would not be fully realized until the keyboard concertos of Mozart

and Beethoven.

 

Copyright by
ROBERT JAMES GALLOWAY
1996

To Diane,
with love and gratitude

 

 

ACKNOWLEDGEMENTS

A number of individuals have given considerable assistance to this project and
thus deserve a word of sincere appreciation. First and foremost are the members of my
guidance committee at Michigan State University: Drs. Dale Bonge, Conrad
Donakowski, and Charles Ruggiero; the outside reader Dr. Richard Hall from the
Department of Philosophy; and most especially Dr. Rosalie Schellhous, who not only
chaired the committee but also offered considerable advice and encouragement which
proved invaluable to the study. With patience and genuine interest, she happily critiqued
and proofread multiple drafts and recensions, and offered considerable help with the
organization and form of the final document. In addition, I am indebted to Dr.
Christoph Wolff of Harvard University for advice in defining the scope of the study and
affirming its value, as well as periodically answering questions and assisting in the
acquisition of research materials.

The resources of several fine library collections and the expertise of their
personnel have been equally indispensable to the progress of this study. The assistance
of the Staatsbibliothek zu Berlin--Preussischer Kulturbesitz, Musikabteilung mit
Mendelssohn-Archiv, in obtaining a microfilm reproduction of the autograph of Bach’s
Concertos for Solo Harpsichord and Strings, BWV 1052-1058, is gratefully

acknowledged. The resources of the Sibley Library at the Eastman School of Music in

i

 

 

Rochester, NY, and of the Music Library at Harvard University, Cambridge, Mass., are
equally recognized. Additionally, the library staff at Houghton College has been most
gracious in providing long-term loans of materials and assistance with numerous
interlibrary loan requests.

The support of two sabbatical leaves from Houghton College, as well as a
University Fellowship from Michigan State University in support of dissertation
research, are also acknowledged with gratitude. Without such support, this project could
not have been realized.

Most importantly, words fail to express the debt of gratitude owed to my wife
and family for their unwavering support, patience, and love through what at times has
seemed an endless enterprise. Their encouragement, personal sacrifice, and cheerful
willingness to abide the disruptions of family life attendant to the pursuit of a doctoral

degree have been deeply appreciated.

vi

 

 

 

Preface

The subject of this thesis arose as a natural corollary to an earlier study of Bach’s
concertos for solo keyboard undertaken some years ago at a National Endowment for the
Humanities summer seminar at Harvard University. As a keyboard player, my interest
in these works was piqued by the prospect of discovering new repertory analogous to the
more famous Italian Concerto, but also by questions which had been raised by scholars
whether such works had served as Bach’s avenue of acquaintance with the Italian
concerto form. Since all 16 concertos for solo harpsichord were also transcriptions of
instrumental concertos by various Italian masters, they offered a seemingly unique
opportunity to explore Bach’s compositional process of adaptation when recasting a
work.

While these earlier concertos dating to Weimar occupy no part of the present
study, they nonetheless provided a springboard for an examination of the later Leipzig-
period concertos for harpsichord and strings, works which are of significance to the
beginnings of the clavier concerto. The ideas for such a study became more fully
crystalized via a discussion with Christoph Wolff, who had directed the Harvard Bach
seminar and is co-cditor of the Bach Jahrbuch. Dr. Wolff endorsed and encouraged the
proposed study, and also suggested examing the obbligato sonatas as well for possible
parallels in the treatment of the harpsichord. 1 am indebted to him for this advice, as

well as for answering occasional questions during the course of the study.

vii

A

 

 

Wherever possible, the Neue Bach Ausgabe has been used as the edition of
choice, and the critical notes accompanying the NBA have been invaluable.
Unfortunately, the seven cembalo concertos which form the centerpiece of the study,
BWV 1052-1058, have not yet appeared in the NBA, and so the older Bach Gesellschaft
edition has served as score for study and excerpted examples for these concertos. Unless
otherwise noted, examples presented in the text are taken from the NBA, excepting the
cembalo concertos BWV 1052-1058 and the fourth Brandenburg concerto, BWV 1049,
which are excerpted from the BC.

In the course of the work, it became apparent that a fuller grasp of Bach’s
transcription process required examination of the autograph manuscripts, since the latter
not only contained signs of multiple layers of compositional activity but also differed in
details from the reading in the Gesellschaft edition. Accordingly, a copy of the
autograph manuscript (Mus. ms. Bach P 234) on microfilm was obtained from the
Deutsche Staatsbibliothek zu Berlin-Preussischer Kulturbesitz, which proved invaluable
to the work on the concertos. A published faksimile of the double concerto in C minor,
BWV 1062, and the A major flute sonata, BWV 1032, has been similarly valuable. A
perpetual difficulty in all Bach research is the lack of original and reliable sources, and
the present group of works are no exception. Not all the works considered in the present
study have extant autographs; those which do have been given greater priority and
authority.

The method of inquiry has centered principally around careful and thorough
comparison between all variant readings of each work in order to discover the nature and

extent of alterations and modifications which Bach made when adapting this music to

viii

A

 

 

solo cembalo. Since a central issue in this transformation is the replacement of the
continue, special attention has been given to the creation of the left hand part, and this
topic is reflected in much of the following discourse. Further, any consideration of
multiple versions of the same work inevitably raises questions about chronology and
influence, and so these issues also figure promiently.

Finally, certain economies have been employed to reduce cumbersome text. To
this end, for the corpus of concertos and sonatas which form the core of this study,
BWV numbers have been omitted, e.g., 1056 represents BWV 1056. BWV numbers are
retained, however, for Bach works outside this group of concertos and sonatas.
Similarly, frequent references to manuscripts have been facilitated by omitting repeated
citations of manuscript Signaturen, e.g., P 612 stands for Staatsbibliothek zu Berlin Mus.
ms. Bach P 612. Further shortened forms of standard literature and sources are listed

in the table of Abbreviations.

 

 

TABLE OF CONTENTS

 

Abbreviations
Introduction

The Background to Bach’s Chamber Style
The Bach Collegium Musicum in Leipzig

PART ONE: The Harpsichord Concertos

I. The Leipzig Concertos
Sources and Chronology

ll. Concertos Transcribed from Extant Models
Accomodation of the Harpsichord Range:

Key Transposition

Idiomatic Adaptation to the Cembalo

III. The Emancipation of the Cembalo from Thorough-Bass
Fashioning the Cembalo Bass
Special Considerations in BWV 1057
Transcription to Two Keyboards: BWV 1062
Added Harmonic Filler
Conclusions

IV. Concertos Without Models: Comparisons with Parallel Versions
Comparisons of Variant Readings

Fashioning the Cembalo Parts
Derivation of the Cembalo Bass

41

48

58

108

111
125
141
145
149

153
154

179
182

 

 

F.

 

V. The Cembalo in Concertos without Extant Models

Performance Questions: Figured Cembalo Parts
Derivation of the Cembalo Bass

Double Cembalo Concertos without Models
Summary of Observations from the Concertos

PART TWO: The Obbligato Sonatas
VI. The Sonatas with Obbligato Cembalo

Chronology, Scope, and Sources

of the Chamber Music
Concerto Elements and Source Questions
The Autograph of 1062/1032: A Unique Source
Chronological Ordering of the Sonatas
Structure and the Cembalo in the Violin Sonatas
The Second Keyboard Question Revisited

VII. The Obbligato Cembalo Part in the Sonatas
Model/1‘ ranscription Comparisons in the Sonatas
Parallel Versions Comparisons
The Character of the Cembalo Parts
Special Features of the Violin Sonatas

Summary and Conclusions

Bibliography

xi

200

205
210
216
226

234

235
244
256
263
274
284

288
291
299
302
313

325

 

 

Abbreviations

Abbreviations used in this study as a convenience to the reader are as follows:

B Dok

BG

BWV

Bach-Dokumente, edited by the Bach-Archiv, Leipzig, as a supplement to
the NBA; 4 vols. (Kassel: Barenreiter; Leipzig: VEB Deutscher Verlag
ffir Musik, 1963-78).

Vol. 1: Werner Neumann and Hans—Joachim Schulze, eds., Schnfista'cke
von der Hand Johann Sebastian Bachs.

Vol. II: Werner Neumann and Hans-Joachim Schulze, eds.,
F remdschnfiliche and gedruckte Dokumente zur Lebensgeschichte Johann
Sebastian Bachs 1685-1 750.

Gesamtausgabe der Bachgesellschaft, Leipzig, 1851—1899 (reprint, Ann
Arbor: J. W. Edwards, 1947).

Bach Jahrbuch, Leipzig, 1904-.

[Bach-Werke-Verzeichnis] Wolfgang Schmieder, Thematisch-
systematisches Verzeichnis der musikalischen Werke von Johann Sebastian
Bach. (Leipzig, 1950).

Kritischer Bericht (critical report) of the NBA.

[Neue Bach Ausgabe] Johann Sebastian Bach: Neue Ausgabe samtlicher
Werke. Edited by the Johann-SebastiamBach—Institut, GOttingen, and the

Bach Archiv, Leipzig. (Kassel: Barenreiter; Leipzig: Deutscher Verlag
fur Musik, 1954- .

xii

 

1:4:

The system of pitch notation is as follows:

/.Q.

 

 

 

 

 

fiL 80 l 4
ha; - - {’31 ,4= " 1"
' 3/ 3/ -e- r
(._.____I
CC_BB C-B C’b C"b' Cn-b"

xiii

 

Introduction

One of the most significant and far-reaching musical developments of the

eighteenth century was reflected in the changing attitude toward the basso continuo or
thorough-bass. Held to be a prerequisite for all music since the beginning of the
seventeenth century, the continuo represented far more than just a performance practice,
for in effect it controlled the very style and texture of virtually all musical composition.
The treble—bass polarity which fully emerged first in early opera and then spread to most
Other musical forms was made possible by the employment of an instrument, usually
keyboard, which filled in the harmony between the outer voices. This standard musical
Practice, which remained immutable through the era, began to weaken in the mid—
eighteenth century, becoming nearly obsolete by the 1770’s and 1780’s in the face of the
growing recognition that its function had become superfluous. It was particularly in
chamber music with smaller instrumentation that the shift away from continuo practice
Was most apparent, and it is this changing function of the cembalo from continuo to solo
itIStrument as evidenced in certain Bach concertos and sonatas which is the principal
focus of the present study.

The group of concertos for harpsichord(s) and strings which Bach transcribed in
Leipzig from earlier concerto settings provides an ideal window for observing this
Process of transformation, since the harpsichord part is derived in part from musical

material which previously had formed part of the continuo. Thus parallel settings, one

.fi

 

.
.-
D'Ii

,..-

”'5.
,, ..-
3 a
. -‘
9 ~ ;
. u
A 'y D
.
. . .
l '\
~..I-.
"'I .

{.5

1'“ .,,

I. .
I.
‘v‘.‘\
-

,.
w».
.; -
.L“.
s“,

.t

 

 

 

I".
1" .‘
. \.

 

5.7‘

 

 

2

with fully realized keyboard parts, one without, illuminate the compositional process.
Further, Bach also wrote a number of chamber sonatas in which an obbligato cembalo
replaces the continuo. These sonatas, interesting in themselves, also invite comparisons
with the above concertos, since in all likelihood the sonatas were written first and
involve similar problems of fashioning a fully realized cembalo part.

Most of the current Bach scholarship on these works has focused on questions of
sources, authenticity, and the original models, rather than on the process of realizing a
viable solo cembalo part. More importantly, these studies have not considered together
these two groups of works which faced similar objectives. The present study attempts
to consider these issues. Of principal interest is the manner in which Bach adapts
musical material to the idiom of the keyboard, with special attention to the formation of

a left hand part, since it is this part which is most closely descended from continuo
Practice. This issue is the principal concern of Chapter 3, but is also considered more
bI'Oadly elsewhere. Since the concertos offer the clearest comparisons between model
and transcription, and hence the best insight into transformational process, they are
Considered at length first, followed by attention to the sonatas with obbligato cembalo

in Chapters 7 and 8.
The Background to Bach’s Chamber Style

The remarkable and extensive musical changes which took place during the
eighteenth century, including the emergence of whole new genres and styles as well as
00mpletely revised performance practices, were mirrored to a considerable extent by

SOCietal transformations of far-reaching import. This period was to be the last in which

A

 

3

music could still be classified according to the three cultural institutions which had long
supported and patronized it: the church, the court, and the theater. These three are
referenced repeatedly by writers of the time when describing musical practice.
Musicians since the seventeenth century had held that music served three functions: to
lead the listener to worship in church, to move his emotions in the theater, and to
entertain him in the chamber. This threefold purpose was reflected by Johann Mattheson
(1681-1764) when he wrote in Der vollkommene Capellmeister (1739) "my true aim [in
here describing the various musical styles at length] is toward the glory and honor of the
Most High in His temple, toward the morality of the stage and toward the pleasure of
the spirit in the chambers and salons. "‘

For many eighteenth century writers, chamber music was associated with the
Physical size of the space where it was performed. Johann Gottfried Walther (1684-
1748) in his Musikalisches Lexicon of 1732 defined chamber music as "that which is
customarily performed in the large rooms of the nobility, "2 while Charles Burney (1726—
1814) as late as 1805 continued to describe chamber music as "compositions for a small
Concert room, a small band and a small audience; opposed to music for the church, the
theatre, or a public concert room."3 The smaller space, and its attendant smaller band
of musicians, was viewed as offering the advantage of greater intimacy, not only

¥

‘Johann Mattheson, Der vollkommene Capellmeister, trans. Ernest C. Harriss (Ann
Arbor: UMI Research Press, 1981), 226.

2Johann Gottfried Walther, Musicalisches Lexicon (Leipzig: Wolfgang Deer, 1732),
130; my translation.

3In Abraham Rees, Universal Dictionary of Arts, Sciences, and Literature (London:
A. sirahcn, 1819), s.v. "Chamber Music."

‘

 

4

between the musicians but also between musicians and auditors. Heinrich Christoph
Koch (1749-1816) in his Musikalisches Lexikon of 1802 compared chamber music to that
for church and theater, and emphasized the refinements which the chamber style
promoted:

Since in chamber music the art was never especially directed towards expressing
religious emotions as in the church, or moral emotions as in opera, but was
intended only to serve the private pleasures of the reigning prince or of the court,
and since, besides, it is only performed in a room and with one player to each
part, the result of all these circumstances was that the older composers took
greater pains with the art products for the chamber, nuanced them more finely,
and assumed on the part of the perfomer greater technical finish than they
considered appropriate in compositions for the church, or for the theater, partly
on account of the size of the building, partly also on account of the larger
number of performers for each part, etc. They imitated (so to speak) the painter
who shades more finely and colors in greater detail a painting intended to be
viewed from close by, than, for example, a ceiling painting which is far from the
eye and in which not only would these nuances be lost, but also the effect of the
whole be weakened. The composers treated the works intended for the chamber
in a similar manner, and thus this type of composition achieved its own
character, which is still indicated by the term chamber style.4

Recognition of such a "chamber style” was fully acknowledged by Mattheson,
Who stated that the essential style of a piece was determined by the music itself
l'egardless of its place of performance, and that chamber music did not lose its identity
When performed in locations other than the chamber. Mattheson’s famous report of
hearing some of Archangelo Corelli’s (1653-1713) chamber sonatas played on the organ
at Vespers in Holland, sometimes with violin concertato, was directed to this very
Point.’ Koch’s suggestion that the chamber style is finer in its execution is also echoed

by Mattheson:

 

 

‘Quoted in Ruth Halle Rowen, Early Chamber Music (New York: Columbia
University Press, 1949), 7.

’Mattheson, Der vollkomene Kapellmeister, § 105, 221.

‘7';

“in“.

I ..
M

“—1
(4—)

 

5
This style in the chamber also requires far more diligence and perfection than
elsewhere, and must have pleasant, clear interior parts which as it were
continually contend for precedence with the upper parts in an agreeable manner.
Slurs, syncopations, arpeggios, alternations between tutti and solo, between
adagio and allegro, etc., are such essential and characteristic things that one for
the most part seeks them in vain in churches and on the stage: because there is
more reliance upon the prominence of the human voices, and the instrumental
style is only used there to improve and to accompany or strengthen; whereas it
clearly asserts superiority in the chamber; indeed, even if the melody should
occasionally suffer a little thereby, it is still embellished, ornamented and
effervescent. That is its distinctiveness.‘S
In Der vollkomene Capellmeister Mattheson delineated five categories of chamber style
in much the same way that he had enumerated multiple styles of church and theater
music. Significant is his recognition that the chamber styles encompass all kinds of
instrumental music--sonatas, concertos, suites, and the like-—thus including what would
now be considered chamber orchestral music as well as chamber music proper. Burney,
in his History, cited cantatas, songs, solos and trios, quartets, concertos and symphonies
0f few parts as all falling within the chamber music category. That distinctions such as
dd chiesa and da camera had become thoroughly blurred by mid-century attests to the
breadth of musical types which was subsumed under the broad designation chamber
Music. The modern dichotomy between concerto and sonata genres is foreign to the
Period, for both existed side by side and equally in the chamber music salons.
Not limited only to the private chambers of the aristocracy, chamber music in the
eighteenth century also reached into the public meeting places of the commoner, either

at home or in other rooms open to the public. Giorgio Antoniotto (1692-1776) reported

that in Italy

 

 

‘Ibid., § 106, 222.

 

aw- -.

LR. L

' a\
l.
.‘ ‘
' o
Fw _
'hv.

 

 

6

all sorts of chamber music was performed in private, and in those particular
assemblies of young gentlemen, where the instrumental music is practised for
pleasure and for practice, every one who is capable of composing, exposes their
musical compositions, which are almost all composed for the violino, flute, or
oboe solo, or in concertos with a principal instrumental part accompanied by the

other instruments.7
Antoniotto’s description is interesting on two accounts: first, its reference to various
types of music, especially works for solo instrument as well as concertos; and second,
the obvious analogies with student gatherings in Germany which provided an important
venue for the hearing of new instrumental pieces. Johann Friedrich Fasch (1688—1758),
a one-time student at the Thomas—Schule where he had been a pupil of Johann Kuhnau
(1660-1722), established in the first decade of the century a collegium musicum in his
IOdgings on Sundays for informal music making. When the gatherings outgrew his small
dwelling, they were moved to Lehmann’s coffee house. Johann Adam Hiller (1728-
1804) later commented that Fasch’s choice of location was well-suited and quite logical,
for such houses had long been a place of music-making in Leipzig.8 The descendent of
Fasch’s collegium was the "other" collegium musicum in Leipzig during Bach’s time.
What is most significant is that the association of these houses with diverse instrumental
Works encompassing suite, sonata, and concerto was already well established when Bach
arrived in Leipzig. The presence of such institutions was of immense importance in

encouraging the composition of some of Bach’s finest works in the chamber style,

including both sonatas and concertos.

\

7Giorgio Antoniotto, L ’Arte Armonica; or, A Treatise on the Composition of Musick
(London, 1760), 109; quoted in Ruth Halle Rowen, Early Chamber Music (New York:
COIumbia University Press, 1949), 7-8.

aJohann Adam Hiller, Lebensbeschreibungen bertihmter Musikgelehrten and
Tonkanstter neurer Zeit (Leipzig: Edition Peters, 1975), 60.

 

7

In one sense, there is a certain irony in Bach’s association with the coffeehouses

of Leipzig. Sebastian Bach, epitomizing the older baroque manner steeped in
counterpoint and thorough-bass, for a few years found himself engaged in the
performance and composition of instrumental works that began to explore new musical
frontiers, most especially a reconsideration of the role of the cembalo, proved through
the simple venue of the Leipzig coffeehouse which itself was a symbol of galant
sensibilities. Commercialism in Germany in the late seventeenth and eighteenth centuries
encouraged an aspiring middle class who were very eager to adopt foreign mannerisms
in order to acquire an appearance of social sophistication. Two of the most important
galant centers were Hamburg and Leipzig, both of which cultivated an active literary
environment as well. The use of coffee and tobacco became associated with galant
Values, and were regarded highly along with reading and card playing for their social
benefits. Hamburg saw the first use of tobacco in the middle of the 17th century and the
fil‘st coffee house in 1680.9 Seen against this background, Bach’s Coffee Cantata (BWV
211) from 1734, undoubtedly presented at Zimmermann’s coffeehouse, gains added
meaning, particularly as a parody of contemporary galant virtues.

The galant was only symptomatic of tremendous social, philosophical, and
Cultural changes which were sweeping western Europe. The decades of the 1720’s to
1740’s were crucial in the transition that ultimately ushered in the new aesthetic of the
age of Haydn and Mozart. Among the altered performance practices alluded to earlier

Which mark this transition, none is more fundamental or far reaching than the changing

\

9David A. Sheldon, "The Galant Syle Revisited and Re—evaluated," Acta
Musicologica 43 (1975): 242-3.

-A

 

o-s

eat

 

 

 

 

 

s
IL_
u

 

8

role of the harpsichord. In the first half of the century, the basso continuo was the
common feature of almost all music. Its advancement of a bass-treble polarity, coupled
with the improvised realization of non-thematic inner parts, impeded the formation of a
true chamber style. Not surprisingly then, composers’ first efforts at reducing or even
eliminating the continuo are found in chamber music, efforts which were abetted by the
ideals of the style galant.

The transition from continuo-dependent to continuo—free compositions was not to
be realized all at once, but steps toward the accomplishment of this end may be found
in certain works of Bach considered in the following chapters. Certainly one phase,
perhaps the earliest, in the disappearance of the continuo was accomplished via
transcriptions in which a former continuo part was transformed into an obbligato cembalo
part. The interchangeability of instruments of like range, as well as the adaptability of
a given composition to other performance situations, not only typified the musical
attitude of the period but also fostered transcriptions of this type, and simultaneously
Offered opportunities for cultivating new chamber music possibilities. The manner and
CXtent by which Bach goes beyond mere transference in adapting a composition to
Obbligato cembalo is of considerable importance in the development of both the future
lCeyboard concerto and the sonata with melody instrument.

It is notable that Bach was not the only composer of the time who undertook to
cOnvert existing works with continuo into settings with obbligato cembalo,lo or who

Sought to compose new chamber music without continuo. Among extant works with

 

 

loExamples of trio sonatas converted into solo sonatas with obbligato cembalo are
f0llndalso in the work of J. G. Graun and F. Benda, as well as C.P.E. Bach. See Hans
Mersmann, Die Kammermusik (Leipzig: Breitkopf and Harte], 1933), III, 162.

 

 

‘

‘.o\

"' ...\
i

.n.~ .

n, ,—
L. .iL l‘

f—fir—WH

\w—o

50-1

sir. i

«J

I”

.071.
.

.jgk‘

'o-e

4-"5.

.,~‘

'1
3'1.

1
r .

 

 

\

_____———_———

I?

 

 

 

9

obbligato cembalo which are contemporaneous with Bach’s compositions in these genres
are the following:

J. G. Graun, 6 Sonatas for Violin and Cembalo (1726-27)

Christoph FOrster, 6 Sonatas for Obb. Cembalo and Violin (1724-27)

Christoph Graupner, Sonatas for Violin (or Flute) and Obb. Cembalo (c. 1740)

Johann Pachelbel, Sonata for Violin and Obb. Cembalo (1690’s)

Johann Matthias Leffloth, 2 "Concertos" for Obb. Cembalo and Violin or Flute

(c. 1730)

Johan Joachim Agrell, Sonata for Cembalo and Violin Obb. (1743)
Other composers who produced sonatas for violin, flute, or viola da gamba with
obbligato cembalo, and whose works are either undated or appeared in the 1750’s and
1760’s are Christoph Schaffrath, Christian Schale, Johann Philipp Kirnberger, Christian
Friedrich and Johann Heinrich Rolle, Johann Gottlieb Goldberg (of ’Goldberg’ Variations
fame), Johann Ludwig Krebs, Christlieb Siegmund Binder, Georg Benda, Johann Philipp
Breidenstein, Johann Jacob Paul Kiiffner, and Bach’s own sons Carl Philipp Emanuel,
Wilhelm Friedemann, and Johann Christoph Friedrich Bach.“ While many of these
composers represent Kleinmeister of the time, their attention to works with obbligato
cembalo nonetheless illustrates a growing contemporary interest in shedding the long-held
dependence on basso continuo.

Of the sonatas and concertos which formed the core of the chamber music
repertory, the concerto offered the greatest opportunities for the development of a fully

realized cembalo part. Although the concerto had been a primary genre of the baroque

almost since its inception, Bach was perhaps the first to see the inherent possibilities in

 

”Alfred Wierichs, Die Sonate fiir obligates Tasteninstrument and Violine bis zum
Beginn der Hochklassik in Deutschland (Kassel: Barenreiter, 1981); and William
Newman, The Sonata in the Baroque Era (Chapel Hill: The University of North Carolina
Press, 1959), 255-300.

A

 

 

 

10

tramforming the harpsichord into the solo instrument. Interplay of thematic material
between harpsichord and strings, or particularly between two harpsichords in the double
concertos, may find a parallel in the intimacy of the sonata, but the intrinsic nature of
the concerto is toward brilliance, a brilliance which encourages both harpsichord parts-—
discant and bass--to fully exploit keyboard technique. Perhaps it was this appeal that led
Bach to return to the writing of concertos in the 1730’s. There may also have been other
stimuli. To begin with, Bach showed renewed interest in solo keyboard music during
the middle Leipzig years, resulting in the Italian Concerto and French Overture (1735),
the Partitas (1727—31), the Chromatic Fantasy and Fugue (rev. 1730), the C minor
Fantasy (1738), and the second book of the Well-Tempered Clavier (1738-42). Some of
this keyboard activity, as well as the writing of the first cembalo concertos, may also
have been inspired by the acquisition of a new harpsichord in 1733. But the attention
to solo keyboard composition does not fully explain Bach’s renewed interest in the
concerto.

In the early 17305, at the very time of Bach’s first years of association with the
Leipzig collegium, Bach renewed contacts with the musical life of Dresden, the court of
the Elector of Saxony. Bach had many musical acquantances in Dresden whom he had
known from as early as 1717; some, including Johann Hasse, Georg Benda, Sylvius
Leopold Weiss, Carl Heinrich Graun, and Jan Dismus Zelenka, visited him in Leipzig.
In 1731 Bach was in Dresden for the premiere performance of Johann Adolf Hasse’s
(1699-1784) opera Cleofide, and while there presented concerts at the Sophienkirche and
at court. Two years later Wilhelm Friedemann (1710-1784) became organist at this very

church, and the Kyrie and Gloria of what was to become the B minor Mass may well

A

 

 

11
have been performed there in 1733. If such a performance took place, Bach personally

may have spent time in Dresden in 1733. But also in Dresden was one of the most well-
known German violinists of the time, Johann Georg Pisendel (1687—1755), who had been
a pupil and protégé of Antonio Vivaldi (1678-1741) in Italy and was particularly noted
for his cultivation of Vivaldi’s style in Dresden. Pisendel had brought back from Italy
many Vivaldi works, including concertos, some copied in Pisendel’s own hand, which
became part of the important Vivaldi manuscripts at Dresden. It is not inconceivable
that the presence in Dresden of so notable a Vivaldi disciple may have prompted Bach’s
renewed attention to the concerto on the Vivaldian model. It is noteworthy that while
Bach had transcribed some Vivaldi concertos for solo harpsichord at Weimar (pre—1717),
he wrote nothing further modeled after Vivaldi until the Concerto in A minor for four-
harpsichords (transcribed from Vivaldi’s Op. 3 no. 10) in the 1730’s.12 Such
hypotheses rest, of course, on conjecture; but in the absence of more substantive
documentation, premise remains valid so long as it is recognized that conclusions drawn
are only as strong as the bases from which they are postulated.

The concertos and sonatas with obbligato cembalo considered in this study
demonstrate the considerable advances made in the treatment of the harpsichord in
chamber music; such advances helped prepare the way for the important style changes
which only became fully established in the second half of the century. Bach’s activities

with the collegium provided a platform for the exploration of these musical innovations,

 

12It is also notable that in two of the Vivaldi concertos in the Dresden collection (PV
226 and 274), the organ appears as an obbligato or concertante part. As will be seen,
this same instrument figures prominently in certain cantatas which parallel several of
Bach’s harpsichord concertos. See Rudolf Eller, "Vivaldi - Dresden - Bach," Beitrage
zurMusikwissenschafi, III (1961): 37-38.

h

3E

{J

5.“:7 _

.w'

r‘.

 

 

 

 

 

 

 

 

 

12

and so it is appropriate to set the stage for this study with a consideration of Bach’s

association with the Leipzig Collegium musicum.

The Bach Collegium Musicum in Leipzig

Music in Leipzig in the first half of the eighteenth century was richly varied and
centered around three institutions which together sustained the artistic life of the city: the
church, the theatre, and the student musical unions, or collegia musica. Two student
collegia were already in existence when Bach arrived in Leipzig in 1723. One of these
had been founded fifteen years earlier in 1708 by Johann Friedrich Fasch (1688-1758),
but was currently led by the Nicolaikirche organist Gottlieb GOrner (1697-1778) (hence
the designation "GOrnerische Collegium"). The second and older collegium had been
founded in 1702 by Georg Phillip Telemann (1681—1767),13 who had arrived in Leipzig
in the previous year to study law. The collegium which he established rose to
prominence rapidly and exercised a considerable influence on the musical life of Leipzig
in the first decade of the century.

Apparently Telemann’s musical talents were manifest soon after his arrival, for
soon he was called upon to compose for the Thomaskirche as well as for the Leipzig
Theatre; he even appeared in productions at the latter himself. In August of 1704,
Telemann gained the position of Organist to the Neukirche. This double association--
director of the collegium and organist at the Neukirche--became standard practice until

1729 when Bach assumed direction of the collegium. It is noteworthy that when

 

”Spitta records the date as 1704. Philipp Spitta, Johann Sebastian Bach iii (New
York: Dover Publications, 1951), 17.

1‘

13

Telemann was installed as organist at the Neukirche, he was also expected "to give his
services at the Thomasschule on occasion-—he was not only to play on the organ, but to
direct all the music--but he must refrain from theatres and give up acting."14

The connection between opera and sacred music, reflected so clearly in the
activities of Telemann, was a strong influence on the musical life of Leipzig. The opera
house, opened in 1693, mounted performances only for short seasons annually, and was
not of the same stature as the operas of Hamburg or Dresden. Nonetheless, Telemann’s
reputation and popularity drew so many student singers to the opera and thus away from
the Thomasschule under Johann Kuhnau (1660-1722) that it had a negative effect on the
quality of the church music. In a memo of March 17, 1709, to the Town Council,
Kuhnau complained of the increasing influence of the opera, which "causes the greatest
mischief, for the better students, as soon as they have acquired, at the cost of infinite
pains to the cantor, sufficient practice, long to find themselves among the
’Operisten."” The intimation here is that some of the same young singers who were
trained for the church music found their way into the opera and, through the personal
affability of Telemann, into the collegium as well. Spitta remarks that music in the

Neukirche was marked by "a lively and operatic style,"“5 and that Telemann’s

 

1“Acts of the Leipzig Council, quoted in Spitta, Bach ii, 182, from Telemann’s
autobiography in Mattheson’s Grosse General Bass Schule (1731), 173.

l51m, Appendix B, p. 304.

"The perceived negative influence of opera on the music of the church is further
reflected in the contract which Bach signed upon his appointment at the Thomasschule,
and which in fact had been prepared for Telemann the year before. Among its points
was the stipulation that the new cantor should "so arrange the music that it may not last
too long, and also in such wise as that it may not be operatic, but incite the hearers to
devotion.“ Ibid., Appendix B, V1, p. 302.

 

14

collegium became the "most important musical institution of Leipzig" for twenty or more
years." Telemann was succeeded by Melchior Hoffmann (1704-1715),18 Johann
Gottfried Vogler (1716-20), and Georg Balthasar Schott (1720-29). After SchOtt’s
appointment to a new post in Gotha in 1729, Bach assumed the direction of this
collegium. Formerly known as the "Telemannische Collegium Musicum, " it soon gained
noted recognition as the "Bachische Collegium Musicum. "

The size of these collegia in the early 18th century is not precisely known, but
if the number of members during Telemann’s or Hoffmann’s associations with collegia
is any indication, then 15-20 Akademikern might be assumed. No doubt most of these
were instrumentalists, since a singers’ quartet of 2 falsettists for soprano and alto parts
plus a tenor and bass was adequate for solo parts in cantatas and as tutti for choral
movements. Spitta stated that under Hoffmann the collegia numbered upwards of sixty
members and met Wednesdays and Fridays from eight to ten p.m. for practice, with
performances only ”on grand occasions, or at Fair time."19 Unfortunately, Spitta gave
no source for this number. It seems unlikely that a collegium would rise to such size
at the time; Spitta even remarked that the regular practisings were held in a coffeehouse

on the market place from which listeners were not excluded. It may be questioned if

 

”ibid. , ii, 205.

u’Hoffmann, an esteemed violinist, made a tour for two years beginning in 1710,
during which time Pisendel took his place in the collegium. Pisendel, himself a member
of the collegium, later became a pupil of Vivaldi and went on to Dresden to make a
considerable name for himself. As already suggested, as Vivaldi’s protégé he may have
had some influence on Bach’s renewed interest in the music of Vivaldi in the 1730’s.

“Spitta, Bach ii, 205.

A

 

15

coffeehouses of the time even could have accomodated so large a gathering. As will be
noted below, even the larger works of Bach known to be associated with collegium
performances at Zimmermann’s coffeehouse do not require so large a number of
performers.

Both of the established collegia met regularly in noted coffeehouses of Leipzig:
GOrner met with his students on Thursdays from 8-10 p.m. in the Schellhafer Haus in
the Klostergasse, while Bach assembled with his collegium in Zimmermann’s
coffeehouse in the Katharinenstrasse. Bach’s group met Fridays from eight to ten in the
evening during winter, but in summer gathered outside in Zimmermann’s garden on the
Grimmischer Stein-Wege from four to six in the afternoon. During the twice-annual
fairs, meetings were held twice weekly, on Wednesdays and Fridays. The consistency
with which these sessions were held is noteworthy: the Friday evening meeting time had
been kept since the time of Hoffmann, if not before, and was still being maintained past
1750, ten years after Bach’s retirement from the collegium.20 The Leipzig
Adresskalender of 1751 announced summer meetings of the collegium in the garden "vor
dem Grimmischen Thore auf der Hintergasse alle Mittwoch Nachmittag von 4-6 Uhr,"
and in winter "in der Catharinstrasse im Oertelschen Hause alle Freitag abend von 8-
10. "1

Bach’s formal association with the "Telemannische" collegium began in 1729,

nearly six years after his arrival in Leipzig. The exact dates of Bach’s relationship with

 

'° Ibid., 205.

21Arnold Schering, Musikgeschichte Leipzigs, iii (Leipzig: F. Kistner & C. F. Siegel,
1941), 147. "Oertelschen Hause" is another name for Zimmermann’s house, evidently
used after Zimmermann died in 1741.

 

16

the collegium are not fully documented, however. That his first term with the collegium
began early in 1729 is known on two accounts: first, the departure for Gotha early in
1729 of the previous director, the Neukirche organist G. B. Schott (1686-1736); and
second, a postscript to Bach’s letter of recommendation for Chr. Gottlieb Wecker dated
20. March 1729. In the latter, Bach wrote a personal note about Schott’s imminent
departure and his own intention to assume the direction of Schott’s collegium.22 The
other dates delineating Bach’s tenure with the collegium are not as clear, but it is
generally accepted that Bach’s direction of the collegium falls into two periods: 1) March
1729 to summer 1737, and 2) October 1739 to sometime in 1741.23

Only limited public notices of performances by the Leipzig collegia were carried
by newspapers or recorded by chroniclers or city archivists. Most commonly, these
notices were limited to events celebrating certain university professors or, more
commonly, special events of the royal family. These latter provided occasion for the so—
called "Extraordinaire" concerts, which were sometimes reported in the newspapers.
The weekly "Ordinaire" concerts, however, found virtually no reference in the press.
Though scant, the meagre announcements of the "Extraordinaire" concerts nonetheless

are significant for the limited information they do provide. Almost invariably these

 

22Werner Neumann, 'Das ’Bachische Collegium Musicum” BJ xlvii (1960): 6. The
letter of testimonial is reproduced in translation in The Bach Reader, 115-16.

2What there is a break of two years is supported by public notices which undeniably
ascribe the direction of the collegium to J. C. G. Gerlach (Bach’s former student from
1723-29). Bach’s return to the collegium in 1739 is attested to in a letter of Sept. 28,
1739, from Johann Elias Bach (1705-55). Here it is stated that Sebastian "will begin the
Collegium musicum this Friday, and will perform some music in the first week of the
Fair for the birthday of His Royal Majesty.” (The Bach Reader, 163). A thorough
examination of questions relating to the dating of Bach’s association with the collegium

may be found in Neumann, "Bachische Collegium," 5—10.

FYI

17

programs for special occasions such as royal birthdays or namedays featured a cantata,
and in many cases the text was printed for the auditors, making identification of the
work possible. A variety of nondescript general titles for the music may be found,
including solenne Musik, solenne serenade, Drama, or Abendmusik. Both Leipzig
collegia considered it their duty to provide extravagant festive music in celebration of
kings and princesses. In eighteenth century Sachsen, three different sovereigns named
Augustus succeeded each other, and so the 3rd of August remained the Nameday
throughout the period. May 12 was the birthday of August II (died 1733), and October
7 was the birthday of his successor August III. Accordingly these three dates are
encountered repeatedly in the public notices, and provided the occasion for some of
Bach’s most compelling secular cantatas.

The following list enumerates Bach’s Huldigungsmusik (Music of homage) from
the period of Bach’s directorship of the collegium and for which some historical record
surviveszz‘

25.8.1731 Geburtstag des Grafen J. F. Flemming
Cantata So kampfet nur, ihr muntern Thane [=BWV Anh. 10]

3.8.1732 Namenstag Augusts 11
Drama per Musica: Es lebe der Kbnig, der Vater im Lande
[=BWV Anh. 11] Text by Picander

3.8.1733 Namenstag Augusts III
Cantata Frohes Volck, vergnugte Sachsen [=BWV Anh. 12]

 

2‘For a complete list, including performances by GOrner’s and Gerlach’s collegia,
see Neumann, "Bachische Collegium," 13-20, and Schering, Musikgeschichte Leipzigs,

123-130.

 

i e

owl

18
5.9.1733 Geburtstag des Kurprinzen
Drama per Musica: Lasst uns sorgen, lasst uns wachen
[=BWV 213]
8.12.1733 Geburtstag der KOnigin
Drama per Musica: T b'net, ihr Pauken! Erschallet, Trompeten!
[=BWV 214]
17.1.1734 KOnigskrOnung in Krakau
Drama per Musica: Blast Lermen, ihr Feinde! versta'rcket die Macht [=BWV
205a]

19.2.1734 Feier des KrOnungsfestes in Leipzig
(possibly a performance of BWV 205a?)

3.8.1734 Namenstag Augusts 111
Drama per Musica: Auf, schmettemde Trine der muntem Trompeten [=BWV
207a]
5.10.1734 Jahrestag der KOnigswahl
Abendmusik der Studierenden auf dem Markt: Preise dein Glilcke,
gesegnetes Sachsen [=BWV 215]

3.8.1735 Namenstag Augusts III
Cantata Was mir behagt, ist nur die muntre Jagd [=BWV 208a]

7.10.1736 Geburtstag Augusts III
Cantata Schleicht, spielende Wellen [=BWV 206]
The notice of October 10, 1736, is the last one prior to 1740 that ties Bach to the
collegium. Public announcements of celebrations of the king’s nameday and birthday
in 1737 through 1739 either avoid any reference to the director, or connect the collegium
with Gerlach or GOrner. The one notable exception is the homage music of April 28,
1738: the cantata Willkommen! 1hr herrschenden Gatter der Erden [=BWV Anh. 13].

This work, however, was performed under the auspices of the University by "Universitat

19
Studirenden." No reference is made to the collegium musicum.” The only definite
record of a Bach collegium performance from the second period (1739-1741) is for
August 3, 1740, the king’s nameday:
Auf Sr. KOnigl. Maj. unsers allergnadigsten Iandes-Herrn hohes Nahmens-Fest
soll heute Nachmittags um 4. Uhr vom Bachischen Collegio Musico im
Zimmermannischen Garten vor dem Grimmischen Thore ein solennes Drama
aufgefiihret werden.“
Unfortunately the notice does not detail the particular work performed.
It is unlikely that the announcements of the collegium performances highlighted
above include the entirety of Bach’s homage music from these years. That there is a
toral absence of notices in the press for the first three years of Bach’s association with
the collegium (1729-1733) is but one hint toward this supposition. The first
Commentaries in the Leipzig Zeitungen concerning the "Extraordinaire" concerts do not
appear until 1733, so that even the important Nameday celebration of August 3, 1732,
is not recognized.27
If the lapse in press coverage of Huldigungsmusik is to be regretted, how much
more the absence of any information about the regular, weekly "Ordinaire" performances
which were such an important part of Leipzig’s musical life for over half a century.

There can be little doubt that the collegium was Bach’s principal venue for chamber

\

2‘Neumann, "Bachische Collegium, " 17-18. The title page of the printed text for this
Wont is reproduced in The Bach Reader, 159.

26Neumann, "Bachische Collegium," 19.

”The scant information that survives about "Extraordinaire" performances during
1729-32 comes from sources other than the Leipzig Zeitungen, e. g., the Breitkopf text
{fruits and Picander’s Ernst-Schertzhcyfte and Satyrische Gedichte. See Neumann,

Bachische Collegium," 13-14.

A

El

20
music in Leipzig. The programs for these performances are unknown, but certainly
included both vocal and instrumental works by a great many contemporary composers.
While descriptions of the "Ordinaire" programs of Bach’s collegium may not survive,
manuscripts copied by Bach’s sons and copyists between 1729 and 1735 strongly hint at
the content of these performances, as the collegium provided the primary avenue for the
presentation of such works. Among performance materials from the Bach circle which
date to the collegium period may be found works by George Frederic Handel (1685-
1759), Pietro Antonio Locatelli (1695-1764), Antonio Vivaldi (1678-1741), Agostino
Steffani (1654-1728), Nicola Porpora (1686-1768), Tomaso Albinoni (1671—1751), and
JOl'tz—mn Bernhard Bach (1676-1749).28 Bach’s name also appears on a list of subscribers
t0 Telemann’s flute quartets published in Paris in 1738.29 In all likelihood Bach’s
Purpose was to acquire more of the latest music available for his collegium.

Alongside such works by Bach’s contemporaries, his own compositions naturally
fOl‘med the central core of the repertory presented in collegium performances, especially
Works for keyboard and instrumental ensemble. It is worth noting that the Clavierubung
I and II, the second book of the Well-Tempered Clavier, and the Goldberg Variations all
aDDeared between 1729 and 1741, the years of Bach’s association with the collegium,
and no doubt most if not all of these works were played at collegium gatherings.
Ensemble works that also find a place here include Overtures, Sinfonias, group and solo
\

”That these works were copied out, no doubt for use in the collegium, also attests

to the regard with which Sebastian held his contemporaries. Andreas Glisckner,
6Ig‘leuerkenntnisse zu Johann Sebastian Bachs Auffiihrungskalender," BJ, lxvii (1981):

2"B Dok, ed. Werner Neumann and Hans-Joachim Schulze (Kassel: Barenreiter,
1969), ii, no. 425. Hereafter, B Dok.

A

21

concertos for violin or cembalo, including transcriptions of foreign concertos, and

sonatas for two or three players.30 Schering speculates that many "premiere"

performances of Bach’s own works must have taken place at the modest Zimmermann

coffeehouse, including violin and clavier concertos, suites and partitas, the Brandenburg
concertos, and many duo sonatas for flute, violin, and gamba.31

Apparently performances of the collegium often included guest musicians, many

of whom may well have been among those who visited Bach on travels through Leipzig.

The following now-famous comment by Emanuel hints at this:

No master of music was apt to pass through this place [Leipzig] without making
my father’s acquaintance and letting himself be heard by him. The greatness that
was my father’s in composition, in organ and clavier playing, was far too well
known for a musician of reputation to let the opportunity slip of making the
closer acquaintance of this great man if it was at all possible.32

\

30Among the instrumental works found in manuscripts copied by CPE Bach are the
A Minor Violin Concerto (BWV 1041), the D Minor Two-Violin Concerto (BWV 1043),
the D Major Overture (BWV 1068), the early version of the D Minor Cembalo Concerto
(BWV 1052a), and the C Major Sonata for Flute and Continuo (BWV 1033). All of
these appear to be datable between 1730 and 1734. GlOckner, "Neuerkenntnisse, ” 71-72.

31Schering, Musikgeschichte Leipzigs, 134. While the Brandenburg concertos clearly
Stem from COthen, it is equally apparent that Bach must have had his own manuscript
copy of them in Leipzig. This is proven by the fact that he transcribed the G Major
c0ncerto (BWV 1049) into a cembalo concerto (BWV 1057) in Leipzig. It is altogether
DOSSible, then, that the Brandenburgs may well have been "premiered" for Leipzig via
the collegium concerts.

32The Bach Reader, 260. While the occasion of a visiting musician "letting himself
_ heard" by Bach could have occurred in the Bach family lodgings, it seems equally
llCely that such a musical encounter could refer to playing with the collegium. From all
unis, Bach’s circle of musical acquaintances was broad: Forkel asserted that Bach
was personally acquainted with Hasse, both Grauns, Telemann, Zelenka, Benda, and "all
most distinguished musicians at that time living in Dresden and Berlin." The Bach

ader, 335.

A

22

No doubt many young musicians, including Bach’s own sons, gained considerable
experience as performers at the collegium concerts. It is altogether likely that it was
Bach’s sons Wilhelm Friedemann and Philipp Emanuel who performed the multiple
cembalo concertos along with their father. From all appearances, the camaraderie of the
collegium not only invited, but encouraged other musicians to participate.” That
Sebastian also was one of the participants in these performances is intimated by Forkel:
In musical parties where quartets or multi-part instrumental pieces were
performed and he was not otherwise occupied, he took pleasure in playing the
viola. If opportunity offered, he sometimes also accompanied a trio or other
work on the cembalo. If he was in a cheerful mood and knew that the composer
of the piece, if present, would not take it amiss, he would either fashion out of

the figured bass a new Trio, or from three separate parts a quartet.34
"Musical parties." Can there be much question that the setting to which Forkel refers
is an occasion at the collegium, particularly given the mention of the personal presence
of the composer? Emanuel also recorded that even to the approach of old age, his father
"Played the violin cleanly and penetratingly, and thus kept the orchestra in better order
t31121.11 he could have done with the harpsichord."35 Where else in Leipzig but at the

cOllegium did Bach have an orchestra to direct, either from the violin stand or, even

more, from the harpsichord?

\
33In the 1736 "Nachricht von den Musicalischen Concerten zu Leipzig" in Mizler’s
Bibliothek appeared the notice ”jedem Musico vergOnnet, sich in diesen musicalischen
Oncerten Offentlich h6ren zu lassen" [any musician is permitted to present himself to
heard publicly in these musical concerts] (My translation.) In B Dok ii, no. 387.

3"The Bach Reader, 334. Forkel received this account via letter directly from
Emanuel; see The Bach Reader, 277, where Emanuel’s letter to Forkel is reproduced and
lll<:ludes this same anecdote.

”Ibid. , 277.

23
As already noted, only the programs for "Extraordinaire" concerts received public
notice; accordingly, the content of "Ordinaire" collegium programs unfortunately remains
unknown. However, it might be suggested that if the instrumentation available to Bach
for collegium programs could be ascertained, then it might be possible to gain some
insight into the repertory presented at Zimmermann’s Coffeehouse. While chamber
works such as sonatas and trios only required a few players, and hence were easily
performable by a modest band of musicians, larger works such as the Brandenburg
concertos or the Overtures, suggested by Schering and others as belonging to the
COIIegium repertoire, obviously required a larger and more diverse band. An
examination of the scoring of known "Extraordinaire" works, especially when compared
t0 Bach’s Thomaskirche cantatas of the same time, sheds interesting light on this
question.

It might be recalled that the conditions under which Bach fulfilled his duties at
the Thomaskirche were far from satisfactory, and complaints with the Consistory had
already begun in the mid to late 1720’s. In the famous "Short but Most Necessary Draft
fOr a Well-appointed Church Music" of 1730, Bach reports the musical conditions at St.
Thomas in some detail, including the number of instrumentalists needed to support the
choir:

Two or three first violins
Two or three second violins
Two first Violas

Two second Violas

Two Violoncellos

One doublebass

Two or three oboes

One or two bassoons
Three trumpets

24

One drum
Two flutes (recorders)
Total: c. 20 musicians.

The enumeration of current players which follows lists only six names, four of which
are Stadtpferlfl’er.36 Bach then goes on to decry this impossible state, adding that the
deficiency has been ameliorated only partly through the use of University students, but
mostly by Thomasschule students, whom Bach must also train in instruments.37
Examination of instrument requirements for the church cantatas of 1728 to 1730
reveals that in almost all cases the scoring demands are fairly light. The following will

illustrate:

Year Cantata Scoring
m3 173 2 Flutes, Strings, cont.
149 3 Oboes, Bassoon, 3 trumpets, strings, cont.
188 Oboe, strings, Violoncello, organ, cont.
197a 2 Flutes, Oboe d’amore, Violoncello, cont.

m2 171 2 Oboes, 3 Trumpets, timpani, strings, com.
156 Oboe, strings, cont.
159 Oboe, strings, cont.
145 Flute, 2 Oboes, Trumpet, strings, com.

1410 12% Viola, cont.
51 Trumpet, strings, cont.
192 2 Flutes, 2 Oboes, strings, cont.

\

36Spitta listed Herren Reiche, Genssmar, Rother, and Gleditsch as Stadtpfeifi‘er, based
on Leipzig archives. However, it would appear that Spitta has crossed names, for Herr
l§Other is listed by Bach as first violinist, while Herr Kornagel, second oboist, is
accounted a ”Kunstgeiger" by Spitta. It would seem then that the latter is the
Stadtpfeiffer, the former the Kunstgeiger. Spitta, Bach, ii, 233, fn. 88.

_ 3"Ibid., ii, 248-249; also The Bach Reader, 121-122. Bach adds that the second
Violins, Violas, celli, and violone have been played by the students, "for lack of more

e1’i’icient performers. "

 

25

Two notable exceptions from 1729 are cantatas 174 and 63; these will be addressed
below. Even if one looks back as far as 1725, the same general outline is present: light
scoring, with preference for oboes over flutes (the use of both oboes and flutes in the
same work is rare). Works which do employ slightly to moderately larger scoring are
almost always associated with "special" days of the church year: for example, Cantata
149 from 1728 for the Feast of St. Michael, or Cantata 171 from 1729 for the Feast of
the Circumcision. In these two and similar cases, some help from University students
was no doubt necessary.
By comparison, the secular homage cantatas which Bach wrote for the collegium
Stand in sharp contrast. Admittedly these cantatas were intended for celebrations of an
overtly festive character. Nonetheless, these works would be altogether unperformable
With the resources Bach had available at the Thomaskirche. The extant Huldigungsmusik

from 1733—36, written during Bach’s first tenure with the collegium, is scored as

fOllows:

26

Year Cantata Event and Scoring

 

1133 213 Birthday of Prince Friedrich Christian.
Oboe d’amore, 2 Oboes, 2 Horns, strings, cont.

214 Birthday of Electress Maria Josepha.
2 Flutes, 2 Oboes, 3 Trumpets, timpani, strings, com.

17 4 205a Coronation of August 111.
2 Flutes, Oboe d’amore, 2 Oboes, 2 Horns, 3 Trumpets, timpani,
Viola d’amore, Viola da gamba, strings, cont.

207a Name Day of August Ill.
2 Flutes, 2 Oboes d’amore, Oboe da caccia, 3 Trumpets, timpani,
strings, cont.

215 Anniversary of election of August 111.
2 Flutes, 2 Oboes, 2 Oboes d’amore, 3 Trumpets, timpani, strings,
cont., plus double choir

1416 206 Birthday and Nameday of August 111.
3 Flutes, 2 Oboes, 2 Oboes d’amore, 3 Trumpets, timpani, strings,
cont.

It should be immediately apparent that this scoring involves completely different
retiources from those which Bach had at his disposal at the Thomasschule. Not only is
the size of the instrumentation of note, but also its diversity: of particular note is the
fl'equent use of both flutes and oboes, including oboe d’amore and oboe da caccia. Also
the larger wind band, plus the frequent performance of this music outside at
Zirnmermann’s Kaffeegarten, would necessitate a larger string ensemble for balance and

sOnority.38 An interesting seating plan showing the layout of the Leipzig Grosse

\

”The summer season, with outdoor performances, apparently ran well into October.
A newspaper notice from 1745 stated that "zum BeschluB der Garten-Lust . . . morgen
den 17. Oct. auf Ansuchen unterschiedener Freunde, in dem Enoch Richterischen Garten
aIlliier . . . nochmals ausserordentliches Concert gehalten werden" [at the close of the
garden entertainment season, tomorrow on the 17th of October at the request of various

A

27
Concert-Gesellschaft in 1746—48, just a few years after Gerlach had assumed the

directorship from Bach, exists in the Leipzig Stadtarchiv, and it shows five first and five
second violins.39 While one cannot assume that Bach’s collegium was this size, it
nonetheless may suggest possible dimensions of Bach’s group.4o
Cantata 174 for Pentecost, June 6, 1729, is also illuminating. Though a church
cantata for St. Thomas’, the scoring constitutes no fewer than 24 parts:
Cantata 174 Ich liebe den Hdchsten von ganzem Gemflte

2 Oboes, Oboe da caccia, Bassoon, 2 Horns,

3 Violins, 3 Violas, 3 Violoncelli, and Cont.
The number of string parts alone probably places this work beyond the resources of the
Thomaskirche. However, since this cantata was written within a few weeks of Bach’s
taking over the collegium, it is most probable that the resources of the collegium were
employed for its performance. Of further note about this particular cantata is the fact

that the opening movement is an instrumental Sinfonia which is a reworking of the first

 

friends, another excellent concert will be held in the Enoch Richter garden] (My
translation.) Neumann, "Bachische Collegium," 24, fn. 45. This means that the annual
Nameday of the king was held outside, as well as birthday celebrations for August III.

39This diagram is reproduced in Christoph Wolff, "Bach’s Leipzig Chamber Music, "
Early Music 13 no. 2 (1985): 167.

”For certain special occasions, especially those held outdoors, a still larger force of
musicians may have been employed. The celebration of the Nameday of August I, May,
1727, resulted in Bach’s cantata Emfernet euch, ihr heitern Sterne! (Anh. 9). Only the
text survives, but accounts of the celebration recount a night-time processional through
the streets of Leipzig with illuminating torches, a Chorus Musicus "alongside other
instruments" and including trumpets and drums. A later description (1745) reports more
than 40 musicians and over 300 students bearing torches. B Dok ii, no. 220. As this
account comes almost 20 years after the event, its veracity may be questioned, but given
the lack of other documentation, it cannot be dismissed altogether.

 

28
movement of the third Brandenburg Concerto (BWV 1048).“1 Thus this large

instrumental ensemble is prominently displayed at the opening of the cantata in a
movement which is very closely associated with instrumental style, i.e., concerto. It is
also interesting that at the very beginning of Bach’s association with the collegium, he
finds himself turning again to concerto models.

The other cantata from 1729 which requires a large orchestral band is Cantata 63
for Christmas:

Cantata 63 Christen, c’z‘tzet diesen Tag
3 Oboes, Bassoon, 4 Trumpets, timpani, strings, Violoncello,
organ, cont.

While there is no Sinfonia as such to open the work, the first chorus commences with
a 33-bar introduction for the full instrumental ensemble. The scoring again suggests that
this work also demanded the resources of the collegium. It is not by accident that these
larger performing forces are called upon for the special days of the church year.

There are also three homage pieces which date from before 1729, i.e., before
Bach’s assumption of the collegium. The first two of these were written in honor of
esteemed professors at the University, while the third is a Trauerode for Queen

Christiane Eberhardine:

 

“The unusual triple division of each string group owes its provenance to the third
Brandenburg, which is scored in exactly the same way (but for strings only).

29

BWV 205: Dramma per Musica for Nameday of Prof. Aug. Friedrich Muller
(3.8.1725)

2 Flutes

2 Oboes
Oboe d’amore

2 Horns

3 Trombones
Timpani
Violins I-lI
Viola
Viola d’amore
Viola da gamba
Continuo

BWV 207: Dramma per Musica for Prof. Dr. Gottlieb Kortte
(11.12.1726)

2 Flutes
2 Oboes d’amore
Oboe da caccia
3 Trumpets
Timpani
Violin I-II
Viola
Continuo

BWV 198: Trauerode (17.10.1727)

2 Flutes
2 Oboes d’amore
2 Lutes
Violin I-Il
Viola
2 Viola da gamba
Continuo (organ and cembalo)42

All three of these works were performed with University students, some of whom must

have been members of the collegium. These works illustrate that even before 1729,

 

42Sicul’s chronicle of the performance of this Trauerode reports that Bach himself
played the harpsichord, and that the instrumentation also included organ and both
recorders and transverse flutes. B Dok ii, no. 232.

30

Bach had begun to establish contacts with some of the University students, who in turn
may have assisted in performances of church cantatas requiring larger scoring. For
example, Cantata 52 (from 1726) includes six wind parts, and opens with a Sinfonia
which uses the opening movement of the first Brandenburg Concerto (BWV 1048). This
circumstance raises the question whether Bach had any role with the collegium prior to
1729. That Bach wrote some chamber works before 1729,“, as well as large sinfonias
to several cantatas between 1726 and 1729, increases such speculations.“

If one assembles a composite instrumentation from Bach’s Huldigungsmusik plus

cantatas 174 and 63 of 1729, the following available instrumentation emergesz“5

 

”The composition of chamber works between 1723 and 1729 does not, however,
necessarily mean collegium performance, though that is a possibility. In his letter to
Georg Erdmann in October, 1730, Bach stated that ”I can already form an ensemble both
vocaliter and instrumentaliter within my family. " The Bach Reader, 126. No doubt the
family circle, sometimes augmented by visitors, did provide a venue for playing chamber
mus1c.

“There is an interesting reference in the Leipzig University chronicle upon Bach’s
first performance of music in the Nicolaikirche, May 30, 1723, which terms Bach ”der
neue Cantor u. Colleggii Musici Direct.” B Dok ii, no. 139. The title here ”Director
of the Collegium Musicum,” is usually taken to be an error which should have read
"Chori Musici.” See The Bach Reader, 93, and Spitta, Bach, ii, 215 fn. 64. There is
also an account by Heinrich Nikolaus Gerber who reported that he heard many concerts
under Bach’s direction while attending Leipzig University in 1724. Spitta concludes that
this can only refer to academic performances for the University, but there is nothing in
Gerber’s report to decisively limit these performances to a particular venue.

“Ji'irgen Eppelsheim claims that after 1730, a number of instruments disappear from
Bach’s instrumental palette, including the viola d’amore, viola da gamba, and recorder.
This is not accurate, as the foregoing discussion attests. Viola d’amore and viola da
gamba are called for in Cantata 205a (1734), and two recorders in Cantata 208a ( 1735)
and BWV 1057 (0. 1736-38). Jiirgen Eppelsheim, "Beobachtungen am Instrumentarium
und Orchester Bachscher Kompositionen aus den beiden letzten Lebensjahrzehnten," J.
S. Bachs Spatwerk and dessen Untfeld, ed. Chr. Wolff (Kassel: Barenreiter, 1988), 77-
78.

31

2-3 Flutes (2 Recorders)
2 Oboes
2 Oboes d’amore
Oboe da caccia
3-4 Trumpets
2 Horns
Bassoon
Timpani
Violins (l-II-III)
Violas (l-II-Ill)
Violoncello (I-Il-III)
Viola d’amore
Viola da gamba
Continuo (including Violone)
Obviously not all of these were used at the same time, and indeed some players doubled
on certain instruments, as oboists might well play oboe d’amore or oboe da caccia.
Interestingly enough, all of the works cited above require 14-16 players, excepting BWV
205a which requires 18 instrumentalists.46 It will be remembered that the size of the
collegium, based on that at the time of Telemann and Hoffmann, was conjectured to be
about 15-20 members, including a vocal quartet. This agrees very favorably with the
performer requirements of these works, especially given the possibility of additional
players drawn occasionally from the Stadtpfeiffer. Furthermore, this size and variety
of instrumentation is quite sufficient to enable performance of the larger ensemble works
(Brandenburg Concerti, Orchestral Overtures) suggested by Schering and others as
belonging to the collegium’s repertory.
The players who participated in Bach’s collegium mostly were students from the

University, some of whom no doubt were quite capable musicians, or visiting virtuosi.

 

46This number assumes one player per string part, which is the way the Brandenburg
Concerti were performed in Cothen and, in all likelihood, in Leipzig. Additional string
players, especially violinists, would increase the number accordingly.

32

These latter no doubt provided Bach an opportunity to compose works without
performance limitations for the enrichment and enjoyment of both performer and listener.
An interesting and significant notice which touches on this appeared in the Leipziger
Lokalblatt Nachricht auch Frag and Anzeiger on June 16, 1733, announcing the
resumption of collegium concerts after the official period of mourning for the death of
August II:

The Collegium musicum may again continue. Therefore, tomorrow, Wednesday

the 17. June in Zimmermann’s garden on the Grimmischen Stein-Wege the Bach

collegium musicum begins with a splendid concert at 4 p.m. in the afternoon, and

will continue weekly, moreover with a new clavier, the equal of which all here

have not yet heard, and which will please both music lovers as well as the

virtuosos who appear here.47
First, the reference to ”virtuosos” appearing with the collegium suggests that the level
of music-making had a reputation for being of high standard, and that this was the
expectation, not the exception. Second, the mention of ”music lovers" ("Liebhaber”)
implies that those who came to listen did so out of genuine interest and appreciation.
Mizler’s Neu ero'fi'nete musikalische Bibliothek of 1736 contained a notice about the
Leipzig collegia, noting that the musicians who performed there were mostly students,
"and there are always good musicians among them, so that sometimes they become, as

is known, famous virtuosos. " Concerning the listeners at such gatherings, the notice

went on to add that ”most often there are such listeners as know how to judge the

 

‘7. . . die Collegia Musica nunmehro wiederum continuiret werden mogen; Als soll
morgen, Mittwochs, als den 17. Junii c. 3. im Zimmermannischen Garten aug dem
Grimmischen Stein-Wege von dem Bachischen Collegio Musico Nachmit. von 4 Uhr der
Anfang mit einem schonen Concert gemachet, und wdchentlich damit continuiret werden,
dabey ein neuer Clavicymbel, dergleichen allhier noch nicht gehoret worden, und werden
auch sich die Liebhaber der Music, wie auch die Virtuosen hierzu einzustellen belieben.
(My translation). B Dok ii, no. 331.

33

qualities of an able musician. "‘8 Most significant is the reference to a new harpsichord
which surpasses anything previously heard and which promises to please the virtuoso.
Nothing else is known about this instrument or its provenance, but this statement
strongly suggests that the harpsichord’s use was not limited merely to continuo
realization but rather was utilized prominently in solo and ensemble music. While large
works quite probably gained a hearing in the collegium’s concerts, chamber music for
one to three or four players, and with obbligato harpsichord, in all likelihood occupied
much of the repertory. In addition to works for solo keyboard already cited which
appeared about this time, it is noteworthy that extant sources for the gamba and flute
sonatas with obbligato cembalo, as well as the solo concertos with strings, all stem from
the 1730’s. Clearly the collegium provided a venue not only for the performance of
these works, but also for proving them.49

Given the variety of instruments available to Bach in the collegium, it is
somewhat surprising to find that certain instruments which figure prominently in the
cantatas and larger ensemble works are altogether absent from the smaller chamber
genres. The obituary written by Philipp Emanuel and Johann Friedrich Agricola shortly
after Sebastian’s death includes the first recorded works list. This inventory, though
incomplete, lists as its last item ”a mass of other instrumental pieces of all kinds and for
all kinds of instruments."’° In the context of the works list, this general collective

statement suggests that the quantity of such works was too great to enumerate. The

 

“The Bach Reader, 149; also B Dok ii, no. 387.
”Neumann, "Bachische Collegium,” 25.

50The Bach Reader, 221.

34

reference to "all kinds of instruments" is all the more noteworthy since the extant
chamber music consists entirely of works for violin, flute, or viola da gamba, with either
basso continuo or obbligato harpsichord accompaniment. Furthermore these pieces are
few in number. When one considers the prominence given the oboe and recorder in the
church cantatas and even the Brandenburg Concerti, the absence of small media works
for these instruments becomes all the more curious.51 It is certainly possible that the
quantity of chamber music which Bach actually composed was far larger (and probably
more diverse) than what is now extant, with much of it having been written to
supplement the repertory of the collegium.

Bach’s earlier positions both at Weimar and Cdthen had offered significant
opportunities for composing and performing chamber music. Rather surprisingly,
however, very little source material for chamber music datable to Weimar or Cothen
survives. This has led some scholars to presume that much of the chamber music found
in Leipzig sources was actually written in Cothen but recast anew in Leipzig. The
absence of significant source material from Cdthen, however, may challenge this
presumption.’2 While loss of pre-Leipzig sources may legitimately be attributed to
problems of transmission, the question remains whether some or even much of this

repertory does not really stem from Leipzig originally. This issue will be considered

 

“Christoph Wolff holds that the absence of sonatas for oboe or recorder is "simply
inconceivable,” and that such works may well have existed but been lost due to the
uneven preservation of Bach’s estate through his sons. While the older sons inherited
the vocal works, the younger sons Johann Christoph Friedrich and Johann Christian
received the chamber works, and much of this material has not survived. Wolff,
"Chamber Music,” 166-67.

52 The only source material definitely attributable to Cdthen is the dedication copy
(autograph) of the Brandenburg Concertos sent to Margrave Christian Ludwig.

35

further in chapters VI and VII. The following table lists the instrumental works which,

if based on extant source materials, probably or definitely stem from Leipzig. It will be

noted that all the sonatas with obbligato harpsichord, as well as the concertos for one to

four harpsichords, appear in this lis

Bach’s Leipzig Instrumental Works

 

BWV Work Scoring Date of source
1014-19a Vl. sonatas v1. hps. c. 1725
1021 VI. sonata vl. bc. 1732-35
1023 VI. sonata v]. be. after 1723
1027-29 Gamba sonatas vla. da gamba, hps. 1736-41
1030-33 Flute sonatas fl. hps. 1730-36
1034-35 Flute sonatas fl. be. 1724, 1741
1039 Trio sonata 2 fl. bc. 1736-41
1041 VI. concerto vl. str. bc. . 1730
1043 2 VI. concerto vl. str. bc. . 1730—31
1044 Triple concerto fl. vl. hps. str. bc. 1729-41
1052-58 Harps. concerti hps. str. . 1738-9
1060 2 Harps. concerto 2 hps. str. bc. . 1736
1061 2 Harps. concerto 2 hps. str. bc. 1732-35
1062 2 Harps. concerto 2 hps. str. be. . 1736
1063 3 Harps. concerto 3 hps. str. be. . 1730
1064 3 Harps. concerto 3 hps. str. bc. 1729-41
1065 4 Harps. concerto 4 hps. str. bc. . 1730
1066 Overture, C 2 Ob. bsn. str. bc. . 1725
1067 Overture, b fl. str. be. . 1738—39
1068 Overture, D 3 tpt. timp. 2 Ob. str. be. 1731
1069 Overture, D 3 tpt. timp. 3 Ob. bsn. str. be. 1729-41

It is significant that by far the majority of Bach’s known chamber music is included here.

Given the opportunities which Bach had in C6then (and Weimar) for chamber music, it

is possible that he composed additional chamber works which have disappeared without

 

53List adapted from Wolff, ”Chamber Music," 169.

36

a trace. While some of the Leipzig—source works do appear to stem from earlier models,
from Cothen or even Weimar, it must not be assumed that all the pre-Leipzig repertory
is resident in the Leipzig chamber works, nor that all of the latter is modeled after the
former.

Neumann raises the question whether Bach’s taking over the collegium in 1729
represents a deliberate step away from the role of cantor and toward the office of
Kapellmeister. Such speculation reflects the traditional view of Bach’s activities neatly
divided between instrumental works composed by Bach the Capellmeister (associated
mostly with Cothen) and vocal works composed by Bach the Cantor (primarily at
Leipzig). As can be shown from the table above, Bach continued to write or rework
instrumental pieces in his early years in Leipzig, well before his formal duties with the
collegium began. Whether these were performed at collegium concerts prior to 1729,
or only in the inner Bach circle, is unknown. However, it is nonetheless apparent that
Bach did not devote himself exclusively to church music in the mid-1720’s, and so the
traditional view that Bach’s move to Leipzig coincided with a fundamental shift of
emphasis from instrumental to vocal music is not tenable.

The question to be raised here is how Bach viewed himself in this context. It is

 

noteworthy that while the titles ”Cantor” and sometimes ”Director Musices" appear
frequently in documents pertaining to the early Leipzig years, the title "Capellmeister”
gradually finds usage as well. During the Leipzig tenure, Bach received the title
Hochfiirstliche Sachsen-Weissenfelssischer Capellmeister. The date when this title was
granted is unknown, but probably was not before the death of Prince Leopold of Anhalt-

Cothen (November 19, 1728), and Bach’s first use of the title in his signature occurs on

37

March 20, 1729.54 This, however, is not Bach’s first use of the title "Capellmeister"
in Leipzig.

From 1727 onwards, references to Bach as "Capellmeister” become very frequent
in the contemporary chronicles as noted by Riemer, Sicul, Vogel, Grabner, Baron,
Henrici, Spiess, Gottsched, and others.” In most cases, perhaps surprisingly, Bach is
mentioned under the sole title ”Capellmeister, " and only occasionally is this title coupled
with his other offices.“5 A list of Sacrament communicants from the Archives of the
Evangelische Thomas-Matthai-Gemeinde, Leipzig, shows at least one Bach entry for
virtually every year from 1723-1749. From 1723 through 1738, Bach is termed
”Cantor, " except in 1727 (note the year), where twice his name appears as "CapellMstr. "
More surprisingly, from 1739 to 1749 (21 entries) Bach is labeled solely
'CapellMstr."” The outstanding question is whether Bach himself ever used the title
"Capellmeister" before 1729, even though his vocational duties were at the
Thomaskirche. The evidence clearly indicates that he did.

In 1727, two years before taking up the collegium duties, Bach wrote letters of

recommendation for Christoph Gottleb Wecker and Friedrich Gottlieb Wild. In both

 

“B Dok ii, no. 327, fn. II. The title reads "Capellmeister to the Prince of Saxe—
Weissenfels as also to the Prince of Anhalt—Cothen, Director Chori Musices Lipsiensis
and Cantor at St. Thomas’s here." B Dok i, no. 60. Also in The Bach Reader, 116.
It is coincidental that Bach also assumed the collegium in March, 1729.

”See B Dok ii, nos. 230, 232, 233, 234, 235, 236, 238, 240, 243, 244, 248, 249,
250, 255, 257, 258, 259, 260, 262.

56For example, ”Herr Johann Sebastian Bach, hochfiirstlich Anhalt-Cothenischer
Capellmeister, wie auch Director Musices und Cantor zu Leipzig." Ibid., ii, no. 255.

5"Ibid., ii, no. 162.

38
cases, Bach signed these ”Joh: Sebast: Bach. Hochfiirstlich Anhalt-Cothenischer

Capellmeister, auch Director Chori Musici Lipsiensis” (to the letter for C. G. Wecker,
he added also "u. den Schulen zu S. Thomae Canton").58 It is worth noting that Bach
places ”Capellmeister” first in order, perhaps suggesting its priority in his own mind.
Still earlier, however, appeared several printed title pages for Cantatas BWV 194, BWV
Anh. 14, and the Clavieriibung l, for which the following ascriptions appear:
Orgelweihkantate BWV 194 (Nov. 2, 1723):
'von Johann Sebastian Bachen, Hochfiirstl. Anhalt—Cothenischen CapellMeister
auch Directore Chori Musici Lipsiensis, und Cantore der Schulen zu St.
Thomas.”9

Trauungskantate BWV Anh. 14 (Feb. 12, 1725):

”von Johann Sebastian Bachen, H. A. C. CapelI-Meister, auch Directore Chori
Musici Lipsiensis und Cantore der Schulen zu St. Thomae.”°°

Both of these title pages were published by the same printer, Immanuel Tietzen, but

obviously the copy for the type setting had to be furnished by Bach. Note that the first

citation above comes only six months after Bach entered service as cantor at St. Thomas.
Partita I from Clavieriibung I (1726):

”by Johann Sebastian Bach / Actual Capellmeister to His Highness the Prince of
Anhalt-Cothen and Director Chori Musici Lipsiensis.M51

In 1731 the six Partitas were published together as Opus 1 with virtually the same title,

except that since Prince Leopold had died on November 19, 1728, Bach deleted this

 

”Ibid., i, nos. 18 and 52.
59Ibid., ii, no. 164.
”Ibid., ii, no. 186.

“The Bach Reader, 105.

.
cw“ ‘3

5‘5

*0

u‘ f;

s. in.

55-3.
I
.8..-

39

prince’s name and replaced it with his more recently obtained title ”Actual Capellmeister
to the Court of Saxe-Weissenfels." Clearly the title ”Capellmeister” was more than a
matter of tradition or custom to Bach, but was intentionally retained, perhaps to remind
those around him that his artistic versatility was not limited only to sacred music.
Earlier the question was raised if Bach had any role in the collegium prior to 1729. To
the scant bits of information that might suggest that he did, Bach’s insistent retention of

the title ”Capellmeister” adds further cause for speculation.

Summary
From the foregoing discussion, the following conclusions may be drawn:

1) While Bach’s duties in Leipzig were formally attached to the service of St. Thomas
Church and School, as well as to the St. Nicholas Church, it is apparent that Bach was
very much involved in the composition and performance of chamber music, for his own
family circle and, more especially, for the collegium musicum. That Bach regarded
instrumental music as an important part of his work as a composer throughout his life
is confirmed by his maintainance of the title Capellmeister during the Leipzig years.

2) The regular meeting schedule of the collegium demanded that a significant quantity
of repertory be readily available, which Bach satisfied with the most current works of
his esteemed contemporaries, most significantly augmented by his own works. The time
demands which this placed on him no doubt resulted in the reworking and transcribing
of some earlier compositions, including works which pre-date Leipzig. However, there
remain a good quantity of instrumental pieces which from all appearances stem from

Leipzig, and in many cases have the collegium as their raison d ’etre.

4O

3) The periods of Bach’s formal association with the collegium musicum are March
1729 to 1737, and October 1739 to at least 1741. There remains the possibility that he
also had some role with the collegium prior to 1729; however this may be, pre-1729
contact and performance with University students for ceremonial and occasional music
is certain.

4) The level of music making within the collegium can generally be assumed to have
been of a high caliber which appealed to virtuoso performers as well as knowledgeable
listeners. The reknown of both Bach and the Bachische Collegium attracted many
visitors, including noted musicians from Dresden and elsewhere who, while passing
through Leipzig, did not fail to call on Bach. Among those who performed in the
collegium were Bach himself, and, in all likelihood, his older musical sons.

5) While the programs of the ”Ordinaire" concerts of the collegium remain unknown,
some idea at least of the available instrumentation may be gained from examination of
the homage music written for special occasions. This instrumentation includes: pairs of
horns, oboes, oboes d’amore; 2—3 flutes (recorders), 3-4 trumpets, bassoon, timpani, and
strings. The last comprises first and second violins, violas, and, when called for, violas
da gamba, violas d’amore, Violoncelli, and continuo.

6) The number of players for upper strings can only be surmised, but probably ranged
from one per part upwards to three or more. That some of the string players were
highly capable is suggested by the music for solo violin (Sonatas and Partitas, as well
as Sonatas with obbligato cembalo) and solo cello, for which some source materials are

traceable to the same collegium years.

‘$,-

b.‘

It 8

H‘

xi:

PART I: THE HARPSICHORD CONCERTOS

 

I.

The Leipzig Concertos

Despite the considerable size of Bach’s creative output, the number of concertos
which he wrote is relatively small: the six Brandenburg concertos, three violin concertos,
a triple concerto for flute, violin and clavier, and thirteen concertos for one to four
claviers and orchestra. Undoubtedly there once existed additional concertos, as is
suggested by the fragment of another clavier concerto (BWV 1059) and the lost models
for BWV 1052, 1053, 1055, 1056, 1060, 1063, and 1064, some of which may have been
scored for instruments not featured in the extant works (oboe, for example).
Furthermore, the number of Bach’s compositions which were strongly influenced by the
concerto, or which were concertos in all but name, is considerably larger. Clearly the
principles of concerto design appealed to Bach and were incorporated into a wide variety
of works, including suites, large choral movements, toccatas, sonatas, and the like.

The concertos proper were written over a span of many years of Bach’s creative
life and may be divided into four groups or phases:

1) Transcriptions of concertos by Italian and Italian-inspired composers, set for

41

42

harpsichord or organ solo. These comprise approximately 22 works, written
mostly in Weimar.l
2) Original concerto compositions for various instruments, including the six
Brandenburgs and the violin concertos. This period extended at least through the
Cothen years, and may have begun as early as Weimar.
3) In the early Leipzig years, Bach wrote a number of cantata ainfonias which
betray concerto characteristics and appear to be reworkings of still earlier
concerto movements that pre-date Leipzig.
4) Between 1729 and c. 1740, Bach wrote the concertos for one to four claviers
with strings, largely to satisfy the repertoire demands of the Collegium Musicum.
Almost all of these works represent transcriptions of Bach’s own concertos for
melody instrument and strings.
In addition to these works are others which, while not entitled "Concerto," display all
the characteristics associated with the genre. Earlier examples of such works include the
third English Suite (BWV 808), the Prelude and Fugue in A Minor (BWV 894),2 and
the G Major Toccata (BWV 916). The first movement of the latter is in all respects a
real concerto for solo keyboard; a copy by Gerber even carried the title ”Concerto seur

Toccata pour le Clavecin. " Datable to 1706 to 1709, it represents one of Bach’s earliest,

 

1An excellent discussion of these works may be found in Hans-Joachim Schulze’s
"Johann Sebastian Bachs Konzertbearbeitungen nach Vivaldi und anderen -- Studien oder
Auftragswerke?", Deutsches Jahrbuch der Musikwissenschaft, xviii (1978): 80ff.
Schulze convincingly places these concerto transcriptions between July, 1713 and July,
1714.

2BWV 894 later served as the model for the outer movements of the Concerto in A
Minor for Flute, Violin, and Harpsichord, BWV 1044.

43

if not the earliest, extant concerto-type works for keyboard.

While these concerto-inspired works may well have had some influence on the
formation of Bach’s keyboard concerto style, it is the works with ripieno strings which
are most important to the development of the keyboard concerto. A number of scholars
have ascribed to Bach the singular honor of creating the keyboard concerto, but with
different works cited to support the claim. Besseler saw the Fifth Brandenburg Concerto
(BWV 1050) as ”revolutionary and apparently without precedent in the introduction of
the clavier as solo instrument, " proving Bach ”der Schopfer des Klavierkonzerts. "3 Karl
Geiringer also held the same work to be pivotal, "the first original cembalo concerto and
a turning point in the history of music."4 While the cembalo certainly carries the
greatest weight among the soloists in this work, the solo duties are shared nonetheless
I with flute and violin, and so, strictly speaking, the fifth Brandenburg is not a true
keyboard concerto. Karl Engel perhaps more correctly traces the beginnings of the
genre to the transcriptions of concertos written originally for melody instruments (BWV
1052-1065), by which Bach "called the clavier concerto into life."5

In comparison with the earlier Weimar transcriptions for solo keyboard, the later
concertos with strings appear to represent something altogether novel for the time. The

practice of arranging concertos, whether one’s own or another’s, for keyboard instrument

 

3Heinrich Besseler, NBA KB VII/2, 27f.
4Karl Geiringer, Die Musikerfamilie Bach, (Munich: C. H. Beck, 1958), 313.

5"Durch die Ubertragung von Konzerten fiir andere lnstrumenten auf das Klavier hat
Bach das Klavier-Konzert ins Leben gerufen. " (”By transcribing concertos for other
instruments to the clavier, Bach called the clavier concerto into life." My translation.)
Karl Engel, 'Konzert," in MGG V11 (1958), col. 1572.

44
alone was not unique to Bach. Bach’s Weimar colleague and cousin Johann Gottfried
Walther had arranged a number of concertos by Italian composers for organ, and this
practice was evidently favored by organists of the time. Mattheson, writing in 1717,
makes reference to this practice as a commonplace when reporting about the playing of
the famous blind Amsterdam organist de Graue, ”who knew by heart all the newest
Italian concertos, sonatas, etc. in three and four parts, and performed them with
extraordinary clarity on his beautiful organ in my presence."6 Even newly composed
works like Bach’s Italian Concerto, though not common, were not unknown.7 But the
clavier concerto accompanied with strings apparently was a new genre altogether, one
which did not appear until the 1730’s.
Bach’s Leipzig concertos with ripieno strings are as follows:
For One Clavier and Strings:
D Minor, BWV 1052 and 1052a
E Major, BWV 1053
D Major, BWV 1054 (After BWV 1042)
A Major, BWV 1055

F Minor, BWV 1056
F Major, BWV 1057 (After BWV 1049)

 

6"Dass auch eben diese Species sich auf einem einzigen vollstimmigen Instrumente
/ Z.E. auf der Orgel oder dem Clavier / zur Curiosité tractiren lass / bewieB unter
andern vor einigen Jahren der beriihmte / aber blinde Organiste an der Neuen Kirchen
auf dem Damm zu Amsterdam / Msr. de Graue, welcher alle die neuesten Italianischen
Concerten, Sonaten, &c. mit 3. a 4. Stimmen auswendig wuste / und mit ungemeiner
Sauberkeit auf seiner wunderschonen Orgel in meiner Gegenwart heraus brachte. "
Johann Mattheson, Das beschiitze Orchestre (Hamburg, 1717), 129.

7The Italian Concerto, part of they Clavier— Ubung II, appeared in 1735. According
to Lothar Hoffmann-Erbrecht, C. E. Rolle first composed 6 unaccompanied clavier
concertos in 1716, followed by Christian Pezold of Dresden in 1729 with his Recueil des
25 Concerts pour le Clavecin. Lothar Hoffmann-Erbrecht, ”Johann Sebastian Bach als
Schdpfer des Klavierkonzerts, " Quellenstudien zur Musik -- Wolfgang Schmieder zum 70.
Geburtstag, ed. Kurt Dorfmiiller (Frankfurt: Peters, 1972), 70.

45

G Minor, BWV 1058 (After BWV 1041)
D Minor, BWV 1059 (fragment of nine bars only)

For Two Claviers and Strings:
C Minor, BWV 1060
C Major, BWV 1061
D Minor, BWV 1062 (After BWV 1043)

For Three Claviers and Strings:

D Minor, BWV 1063
D Major, BWV 1064

For Four Claviers and Strings:
A Minor, BWV 1065 (After Vivaldi’s Op. 3 No. 10)

There is general agreement among scholars that these concertos were composed to supply
repertoire for the collegium musicum concerts. As noted in the previous chapter on
Bach’s Collegium, the presence of a new harpsichord beginning in June, 1733, ”the likes
of which all here have not yet heard," may well have given Bach additional incentive to
compose works which gave extra prominence to the harpsichord. Among such works
may well have been not only many of these concertos, but also sonatas for melody
instrument with obbligato harpsichord. All of the concertos listed above, with the
possible exception of one (BWV 1061), are believed to be transcriptions of earlier
concertos for melody instrument(s), by Bach himself or, in the case of the four-clavier
concerto, Vivaldi.

Werner Breig postulates that if one adopts a definition of the clavier concerto
consonant with the later Viennese understanding of the term, then Bach did not write in

the genre. This definition, according to Breig, submits to three qualifications: written

46

for a clavier, with ripieno, and an original composition.8 The last requisite apparently
would delete Bach’s concertos from consideration. However, it must be considered that
while these works are not newly composed, they are, with one exception, Bach’s own,
but now cast in a different mold. Whenever Bach undertook to revise or transcribe his
own compositions, the end result was always a newly fashioned work, heavily indebted
to its model, but sufficiently restructured and newly detailed as to be considered a new
representation of the original work. Approached on these terms, these works require
serious consideration in any study of the genesis of the clavier concerto.

There are two principal and important accomplishments which Bach has realized
in these keyboard concertos and which help to establish the history of the genre: first,
the emancipation of the clavier from its traditional role as Generalbass instrument, and,
second, the working out of the relationship between the harpsichord with its particular
tone characteristics and technique, and the ripieno string group. The historically well-
established baroque practice of always undergirding the ensemble with keyboard continuo
had to give way to a wholly new concept. Although the clavier had long been present
as a member of the ensemble, its traditional function in the continuo group may have
been more an impediment than an inducement for it to assume a soloistic role.

As late as 1753, C. P. E. Bach wrote in his Versuch fiber die wahre Art das

Clavier zu spielen that no piece of music can be well performed without the

 

8Werner Breig, "Johann Sebastian Bach und die Entstehung des Klavierkonzerts, "
Archiv fiir Musikwissenschafi, XXXVI, Heft 1 (1979): 23—24.

47

accompaniment of a keyboard instrument.9 As long as this requirement was held to be
obligatory and immutable, change would be the more difficult. Some scholars have
suggested the possibility that performances of these concertos may have included a
second harpsichord which fulfilled the traditional continuo function. Despite a lack of
definitive evidence, this is a question that has adherents on both sides and to which the
present study will return in later chapters. While the use of a second keyboard for
continuo realization might have been possible, it nonetheless would make performance
conditions more difficult. It is unfortunate that there is no documentation concerning the
exact number of instruments, keyboard or otherwise, which was present at
Zimmermann’s coffeehouse. However, given the presumed limitations of space and the
numbers of likely auditors, the possibility of multiple harpsichords seems less likely.
These conditions also have import to the discussion of concertos for three and four
harpsichords.

However this matter of the number of keyboard instruments may stand, the fact
remains that Bach had already begun to free himself from the necessity of a keyboard
continuo, as evidenced by the chamber sonatas with obbligato cembalo. At least some
if not all of these appear to predate the concertos, and thus they provided a proving
ground as it were for the writing of ensemble music without continuo. The most
singular advance to be gained by the disengagement of the cembalo from its former

continuo role was the freeing of the left hand part, moving it away from merely

 

9"Man kann . . . ohne Begleitung eines Clavierinstruments kein Stiick gut
auffi’rhren. " (”One cannot play any piece of music well without the accompaniment of
a keyboard instrument.” My translation). C. P. E. Bach, Versuch iiber die wahre Art
das Clavier zu spielen, facsimile edition edited by L. Hoffmann-Erbrecht (Leipzig:
Breitkopf and Hartel, 1969), 2.

48

providing the harmonic bass to giving it a more active role in the working out of the
piece. This all-important change was absolutely requisite if the clavier was to gain its
niche as the solo instrument. In concertos, this also meant making the left hand more
brilliant and virtuosic.

The second matter, the relationship of the harpsichord to the ripieno group, is
focused in two considerations: first, the sound medium itself, and second, the technical
adjustments necessary to accomodate what originally was written for violin to an idiom
suitable for harpsichord. In terms of instrumental timbre, the rapid sound decay of the
harpsichord does not balance easily against the sustained tone of the string ripieno, a
circumstance which had to be accommodated. This was not a factor in continuo playing,
where the primary musical interest lay elsewhere. Technical accomodation to a keyboard
instrument is evident in most of these concertos and mainly concerns changes made in
the right hand part which typically is derived from the solo violin part of the original
model. Despite Bach’s attention to such details, there still are many passages where the

original violinistic figuration is retained in the keyboard writing.

Sources and Chronology

The dating and chronology of these keyboard concertos has been given much
study by many scholars. Fortunately, a number of the concertos survive in original
sources which allow fairly precise dating. The concertos which appear to stem from the
first collegium period (1729-37) are as follows:

1. BWV 1065. Three parts, probably belonging to the original set of parts, for

the Concerto in A Minor for four claviers, BWV 1065, survive in $83 Mus. ms.

49

Bach St 378.10 These parts, in an unidentified hand, are for claviers II, III, and
IV, and bear watermarks which also appear in Bach autographs between 1727 and
1731.“

2. BWV 1052a. This concerto in all likelihood is an earlier version of BWV
1052, though the relationship between the two concertos has been debated. BWV
1052a is preserved in Mus. ms. St 350 in the hand of Carl Philipp Emanuel
Bach, who may very well also have been the author of the arrangement. Siegele
offers compelling observations to support his thesis that 1052, 1052a, and the
Sinfonia of Cantata 146, which also uses the same music, are independent of each
other and are all based on a lost violin concerto. Noting a seeming relationship
with a motive in a trio sonata by Philipp Emanuel, Siegele also posits a possible
date of around 1731.12 On the basis of handwriting, Wilfried Fischer places
BWV 1052a around 1732-34.13 In either case, this version of the work falls
sometime during the early collegium years; even if arranged by Philipp Emanuel,
perhaps as a student work, there remains the strong likelihood that it was
performed as part of the collegium repertory.

3. BWV 1061. The autograph clavier parts for this concerto for two claviers,

 

loUnless otherwise indicated, all manuscript sources cited are from the
Staatsbibliothek zu Berlin-Preussischer Kulturbesitz; subsequent reference to these
manuscript items will not repeat the library designation.

"NBA KB VII/6, 78.

12Ulrich Siegele, Kompositionsweise und Bearbeitungstechnik in der
Instrumentalmusik Johann Sebastian Bachs (Neuhausen-Stuttgart: Hanssler-Verlag,
1975), 106.108.

”NBA KB VII/7, 36f.

50

in the hand of Anna Magdalena and Sebastian, have the familiar "MA, grosse
form” watermark which is assignable to 1732-1735.“

4. BWV 1063, 1064. No autographs survive for these two concertos for three
claviers (in D Minor and C Major respectiveIY); the earliest source for both is
found in score copies by Agricola which date to 1739-41. Many questions
surround these concertos, including their dating, models, and authenticity.
Assuming them to be genuine Bach works originally, their date of composition
certainly precedes Agricola’s copies. In the Preface to the first edition of BWV
1063, Friedrich Konrad Griepenkerl, based on information given him by
Forkel,15 attributed the appearance of these concertos to the probable
circumstance ”that the father wanted to obtain for both of his oldest sons, W.
Friedemann and C. Ph. Emanuel, the opportunity to instruct them in all kinds of
performance."16 If one accepts this tradition, and it has enjoyed long
acceptance, then the concerto in question, and likely its sister work BWV 1064,
could have appeared no later than about 1733-34, as Wilhelm Friedemann left for
Dresden in 1733, and Philip Emanuel moved to Frankfurt a. d. Oder in 1734.
There is also the possibility that some of these multi-keyboard concerti may have

found a further participant in Bach’s student Johann Ludwig Krebs (1713-1780),

 

l4NBA KB VII/5, 75.

15Griepenkerl, according to his own testimony, was Forkel’s student in Gdttingen in
1806.

‘6'. . . daB der Vater seinen beiden altesten Schnen, W. Friedemann und C. Ph.
Emanuel Bach, Gelegenheit verschaffen wollte, sich in allen Arten des Vortrags
auszubilden." (My translation.) Preface to Griepenkerl’s edition, C. F. Peters, number
258 (Leipzig, 1845), cited in NBA KB VII/6, 26.

51

who, according to Bach’s own testimony, had distinguished himself on the
clavier, violin, and lute, as well as in composition.17

5. BWV 1062. This concerto for two claviers in C minor is an arrangement of
the well-known Concerto in D Minor for two violins, BWV 1043. The
autograph score (Mus. ms. P 612, which also contains the A major flute sonata,
BWV 1032) is datable by its watermarks to the fall of 1736,18 past the time of
the departure of Bach’s sons, but within Bach’s first collegium period. Perhaps
J. Krebs served here again as second clavier player alongside Bach himself. The
fall of 1736 places this work at the same time as Mizler’s announcement of the
collegium concerts, in which the participants are noted as ”chiefly students here,
and there are always good musicians among them, so that sometimes they
become, as is known, famous virtuosos."19

6. BWV 1060. The oldest surviving source for this concerto for two claviers
in C Minor is a parts copy by the Bach student Johann Christoph Altnickol,
preserved in Mus. ms. St 136 dating from the 1740’s. No documentation exists
to indicate when this work first appeared; however, it seems likely that it is close
in time to its sister work BWV 1062, i.e. 1736.20

It is notable that all the foregoing keyboard concertos which appear to stem from

 

l"B Dok i, 139; also in The Bach Reader, 135. Bach’s testimonial for Krebs is dated
24.8.1735.

18NBA KB VII/5, 61, 70.
1”The Bach Reader, 149; B Dok ii, 277-78.
20NBA KB VI/S, 13-14, 36.

52

Bach’s first period of association with the collegium are for multiple harpsichords. This
fact is supported by the presence of Bach’s sons in Leipzig through these years. Stated
another way, after the departure of Wilhelm Friedemann and Philipp Emanuel, the
production of concertos for two or more harpsichords seems to have ceased, and instead
Bach turned to composing concertos featuring only one harpsichord (BWV 1052-58).
This circumstance would suggest that Bach himself may well have been the cembalo
soloist for their performance, particularly if one accepts the likely notion that the multi-
keyboard concertos were intended for Bach and his sons.

There remain, then, the seven concertos for one harpsichord with strings, which
are collected together in Mus. ms. P 234, the important autograph score of these works.
Seemingly the existence of such an authentic primary source should facilitate fairly
precise dating. This is only partially true, however. The principal watermark for P 234
appears only in a limited number of documents which stem from 1738, and Hans
Joachim Schulze has suggested that 1738 might also be accepted as the date for P 234.21
However, since Bach was not associated with the collegium in 1738, the writing of these

seven concertos would lack an apparent purpose.22 Schulze goes on to note that also

 

21These documents consist of a receipt of May 5, 1738, which acknowledges
payment for performance of an evening serenade for the king and his company by
University students on April 27,1738 (The Bach Reader, 160; the piece performed was
the lost cantata Willkommen, ihr herrschended Garter der Erden, BWV Anh. 13), and
the autograph score of Cantata 30, F reue dich, erlo'ste Schar, written in 1738 at the
earliest. See Hans Joachim Schulze, Preface to Konzert D-dur fir Cembalo and
Streichorchester, BWV 1054, Faksimile-Reihe Bachischer Werke und Schriftstiicke, Bd.
11 (Leipzig: VEB Deutscher Verlag fiir Musik, 1972), n.p.

22Perhaps it is for this reason that Siegele posits 1737 as the date for P 234, thus
placing the autograph at the end of Bach’s first term with the Collegium. Siegele,
Kompositionsweise, 116.

53

in May of 1738, Bach made a somewhat extended visit to Dresden, and speculates that
P 234 may have had some connection with this Dresden excursion.23 This hypothesis,
while attractive in some respects, lacks documentation to fully support it.

Manuscript P 234 is unique in the Bach sources: except for the dedication copy
of the Brandenburg concertos, it is the only collection of concertos gathered together into
a corpus by Bach himself. Yet there are cryptic hints in the manuscript that this
"corpus” originally may have consisted of only the first six concertos.24 At the end of
concerto six (BWV 1057), alongside the usual ”fine,” appears "SDGI" (”Soli Deo
Gloria”), while at the beginning of concertos one (BWV 1052) and seven (BWV 1058)
is found "JJ' (”Jesu juva"). Furthermore, the titles of the concertos in P 234 are not
uniform: concertos two through six are entitled ”Concerto a Cembalo certato . . ." and

concerto one is similarly "Concerto a Cembalo concertato . . . . By contrast, the

seventh and fragmentmentary eighth concertos are entitled "Concerto a Cembalo obligato

and ”Concerto a Cembalo solo . . . respectively. Beyond these unusual
graphological features, further hints about the autograph’s original content may also be
implied by the layout of the concertos within the manuscript itself. Concertos one

through six are copied out in such a way that the ending of one concerto and beginning

 

23Hans-Joachim Schulze, "Johann Sebastian Bachs Konzerte - Fragen der
Uberlieferung und Chronologie, " Beitrdge zum Konzertschafi'en Johann Sebastian Bachs,
ed. Peter Ahnsehl, Karl Heller, und Hans-Joachim Schulze (Leipzig: Breitkopf and
Hartel, 1981), 12-13. Schulze further speculates that while at Dresden, Bach may have
had opportunity to perform the Concerto in C Major for two claviers, BWV 1061, with
his son Wilhelm Friedemann, organist at the St. Sophienkirche since 1733.

24Such an original collection of six concertos would agree with Bach’s customary
preference for works grouped in sixes: one need only recall the English and French
Suites, the Partitas, the Toccatas, the Violin Partitas and Sonatas, the cello suites, and,
most especially, the six Brandenburg Concertos.

54

of the next occur within a common fascicle, while concerto seven begins on a wholly

new fascicle. The following table illustrates (Fig. 2-1):25

 

 

 

Fascicle pgs. Concertos
1 1- 8 1
2 9-20 --BWV 1052
3 21-28 24 24
4 29-36 --BWV 1053
5 37-48 38 39
|--Bwv 1054
6 49-56 50 51
|--BWV 1055
7 57-64 62 63
|-—BWV 1056
8 65-76 73 73
9 77-84
--BWV 1057
10 85-86
11 87-94 94
12 95-102 95
|—-BWV 1058
13 103-106 106 106--BWV 1059

 

 

Fig. 1-1. Fascicle/Page layout of Mus. ms. Bach P 234.

This arrangement strongly suggests that the seventh concerto and the fragment of the

eighth (BWV 1958 and 1059) were appended to the manuscript at a later date.26 Based

 

2’Table from Werner Breig, ”Zum Kompositionsprozess in Bachs
Cembalokonzerten, " Johann Sebastian Bach: Spa’twerk and dessen Untfeld, Bericht fiber
das wissenschaftliche Symposium anlasslich der 61. Bachfestes der Neuen
Bachgesellschaft Duisburg, 1986, ed. Ch. Wolff (Kassel: Barenreiter, 1988), 45—46.

2“"Siegele, based on Paul Hirsch’s incorrect information that concertos six and seven
shared a common fascicle, concluded that BWV 1058 and 1059 were not attached later,
but must have been written in by Bach at a later date. The correction of the manuscript

55

on style considerations, Breig even suggests that BWV 1058 and 1059 were actually
written out before BWV 1052-1057, but weren’t added until the time of binding.27
However that may be, the available evidence seems to suggest that the first six concertos
formed a self-contained corpus. If so, this might strengthen Schulze’s case for a
connection with Dresden, as the preparation of such a score containing a series of six
concertos (Bach’s preferred work-group number) would then reflect on some particular
purpose. Once that purpose had been fulfilled, whether successful or not, additional
works might then have been added.

It must also be recognized, however, that the autograph contains many corrections
and thus is a working score, not a dedication copy anything like the score of the
Brandenburgs presented to the Margrave. It is even possible to suggest that the score
could have been prepared for some purpose in Dresden, which remained unfulfilled, and
the corrections were then made later through the experience of performances given by
the collegium. The speculative possibilities here seem endless, and there is no point in
pressing for unattainable conclusions. Nonetheless, Bach’s sudden interest in the
keyboard concerto, to the extent that he collected together in a single corpus such a
group of works, underlines the significance of this form to the composer at this time.
Even if not prepared for Dresden, Bach’s visit there may have encouraged him to further
experiment with the writing of concertos, inasmuch as a number of the best Italian-

trained musicians were active at Dresden, including Johann Georg Pisendel, one of

 

detail by Fischer and Weiss alters this conclusion. Siegele, Kompositionsweise, 117; W.
Weiss in NBA KB IX/ 1, 51, 57.

27Breig, ”Kompositionsprozess," 46-47.

56
Vivaldi’s protégés who specialized in performing the latest concertos. Perhaps in

response to such inspiration, Bach set about to transcribe certain pre-existent concertos
for keyboard solo with strings on his return to Leipzig.

Details of the construction of P 234, though interesting in themselves, do not
ultimately answer questions about the dating of these concertos. At least for BWV 1052-
1057, however, there remain the following three possible dating scenarios:

1) The first six concertos in P 234 were composed in 1738, and hence were not
intended initially for collegium performance, but rather were for Bach’s own use at
home or for a wider appeal elsewhere, such as Dresden.

2) P 234 was written before 1738 (i.e. 1736-37), and thus in the last years of
Bach’s first collegium period.28 Perhaps additional works were written on the
same paper during this time, but have been lost. The absence of such works,
however, does not prevent the possibility of a pre-1738 date.

3) P 234 was written after 1738, i.e. 1739, for Bach’s second collegium term,
perhaps in response to renewed interest in the concerto as sparked by Bach’s
Dresden visit of May, 1738.

Whatever the chronology of the sources, it nonetheless remains that the solo
concertos found in P 234 appear to have been worked out by Bach in a relatively short
time, and were performed at the ”Ordinaire" concerts of the collegium between 1737 and
1741. It is also very likely that Bach himself may have served as harpsichordist for

these concertos. It is also possible that Krebs or Johann Friedrich Agricola, another

 

28Both Siegele and Kobayashi place P 234 ”um 1737. " Siegele Kompositionsweise,
117; Yoshitake Kobayashi "Zur Chronologie der Spatwerke Johann Sebastian Bachs
Kompositions- und Auffiihrungstatigkeit von 1736 his 1750," B], 74 (1988): 41.

57

Bach pupil who was in Leipzig from 1738 to 1741 and is known to have played the

harpsichord for performances of the collegium, may have fulfilled this role as well.29

 

29Schulze, Preface to Konzert D-dur.

II.

Concertos Transcribed from Extant Models

Three of the seven complete concertos found in P 234, as well as the Concerto
for two harpsichords in C Minor, BWV 1062, survive in earlier versions, a circumstance
which permits direct comparison between model and arrangement. Accordingly, these

works are of the greatest value to the present study, and are listed here with their

models:
Tmcription M el
Clavier Concerto Violin Concerto
in D Major, BWV 1054 in E Major, BWV 1042
Clavier Concerto Violin Concerto
in G Minor, BWV 1058 in A Minor, BWV 1041
Clavier Concerto Brandenburg Concerto #4
in F Major, BWV 1057 in G Major, BWV 1049
Concerto for Two Claviers Concerto for Two Violins
in C Minor, BWV 1062 in D Minor, BWV 1043

The absence of models for the remaining four concertos in P 234 must not be taken to
indicate that these works were newly composed, however. All seven concertos clearly
appear to have been based on earlier works, an assessment which is supported by the

following observations:

58

59

1) As already noted, at least the first six concertos in P 234, if not all seven, were
most likely written in a short time span. The fact that three of the concertos were
definitely derived by transcription would appear to fit this circumstance, and it seems
likely that the other four concertos may have been similarly produced, given Bach’s
time constraints.

2) For the concertos lacking known models, Wilhelm Rust long ago noted figuration
in the transcribed clavier part typical of violin writing, including widely spaced
arpeggiation and bariolage.1 Such figurations, while usually playable on a keyboard,
nonetheless betray their string heritage.

3) Perhaps most importantly, the autograph manuscript betrays evidence of many
corrections, particularly in the clavier part. Clearly the autograph was a working
copy, and corrections are visible in concertos both with and without known models,
strongly suggesting adaptation from other sources.

For the three concertos with extant models in P234, comparison of model and

transcription reveals that Bach generally followed a relatively direct and consistent

method of transcription. Ripieno string parts are taken over with almost no changes

while the keyboard parts are often treated with greater freedom. The original solo violin

part most commonly is placed in the right hand of the cembalo, and while the transferral

is often quite literal and direct, embellishments frequently are added. The result is an

enhancement of the virtuosic character of the writing and better accommodation to a

keyboard idiom. Although the left hand part of the cembalo is derived closely from the

continua, it often departs significantly from the original continuo bass, and it is this

 

lWilhelm Rust, BG, Bd. 17, xiv, xv.

60

disengagement of the left hand from the continuo bass which is of greatest significance
in giving the cembalo its rightful claim to solo status.

Hints about Bach’s compositional method in adapting the original solo instrument
to the harpsichord may be discovered from the autograph itself. It is evident throughout
the manuscript that the ripieno string parts were copied out first, with the harpsichord
and continuo parts following. This sequence is discernible on two grounds. First, the
amount of lateral space allowed for each measure as delineated by bar lines is dependent
entirely on the upper string parts, with the result that not infrequently the keyboard parts,
especially the discant, must be squeezed into less space. than is really needed. Any

number of examples from P 234 might illustrate, such as the following in Ex. 2-1:

Ex. 2-1. 1054 excerpts from P 234

I, 85-87

 

61

I, 114-115

 

And an extremely obvious example:

I, 33-35

 

Had the more florid harpsichord part even been considered at the same time that the
upper parts were copied, then adequate space might have been provided by increasing
the length of the bars. When copying in the upper parts, three lower staves without bar
lines were left vacant for the keyboard and continuo parts. Only after the upper parts
had been fully copied did Bach go back and write in the keyboard parts, with changes

and embellishments. The bar lines were then extended downward from the upper staves.

62

This copying process leads to a noteworthy conclusion: that Bach evidently had
little concern about possible changes in the strings which the keyboard adaptation might
require. In other words, it appears that Bach intended and expected to transform the
model into a harpsichord concerto almost solely by revisions and modifications of the
keyboard part. In fact the autograph bears this out: while changes, even multiple layers
of change, are apparent in the cembalo part, nowhere do alterations of real substance
appear in the string parts.2 Virtually the only modifications to the latter appear to be
those necessitated by the key transposition, or minor changes of a note here or there
which have nothing to do with the substance of the work. Considering the extent of the
revisions made in the cembalo parts, and especially in the bass, it is all the more
remarkable that no substantive changes in the ripieno strings were deemed necessary or
desirable.

A second ground for concluding that the lower parts were written in after the
ripieno strings is the appearance of the handwriting itself. Although the hand is the same
for both upper and lower staves, in many places the clarity of the writing is poorer in
the lower staves than in the upper,3 which may imply greater hurry in copying down the
material for the harpsichord and continuo. The following excerpt may serve to illustrate

(Ex. 2-2).

 

2Werner Breig notes an instance in the first movement of 1052 where several pitches
in the first and second violins (mm. 142-143) were altered to avoid parallel fifths with
the cembalo discant. The modification is visible in P 234. However, the change is of
little musical consequence, altering neither the harmony nor the motivic content of the
passage. Brieg, "Kompositionsprozess," 34, 36.

3Wilfried Fischer describes the cembalo script as of "konzeptartigen Charakter.”
NBA KB VII/7, 16.

63
Ex. 2—2. 10541, 103-107 P 234

f" , "AL/‘- ’

   

 

‘C

 

0»

Could this more cursory hand intimate that perhaps some passages were proved at the
keyboard, and copied out or altered as they were found acceptable? Such a process
could have taken place either in Bach’s private quarters or in rehearsals with the
collegium. Such a possibility raises additional questions, however, including whether
the harpsichordist played from the score or from a separate part. While the existence
of the autograph score is most important and fortuitous, it is regrettable that further
performance materials, namely the parts, do not survive.‘ If changes to the clavier part
were made in rehearsal or following performance, such alterations would have been
recorded in the cembalo part and not necessarily in the score, unless the harpsichordist
performed from the score. P 234, however, does not betray signs of significant
performance use, although this may be difficult to ascertain, and as already noted there
is no record of the number of performances which these concertos received. The

question of whether or not the harpsichordist performed from the score is difficult if not

 

‘This is true with the exception of 1055, for which some performance materials
survive, and 1057, for which a cembalo part is extant.

64

impossible to determine. On the one hand, the extended passages in P 234 which Bach
wrote out in tablature would seem unnecessary unless the score was used for
performance. For the A Major concerto, BWV 1055, a set of parts (Mus. ms. St 127)
survives for violins, violas, continuo, and a separate part for violone, but no cembalo
part, suggesting that the harpsichordist played from the score. However, a separate part
for cembalo concertante does survive for BWV 1057 (Mus. ms. St 129), as well as for
the early version of the fifth Brandenburg concerto (Mus. ms. St 130). The existence
of keyboard parts for these two works argues against the soloist playing from the score
for performance.’
Despite the seeming lack of special attention to string parts in the autograph,

it must not be assumed that Bach’s method of transcription was mechanical, particularly
in regard to the ripieno parts. The high degree of correspondence between model and
arrangement might prompt the conclusion that the composer paid less careful attention
to the smaller details when reworking a composition. That this is not the case can be
affirmed on several grounds, including the careful adjustments which Bach made to
accomodate the different range of the violin and cembalo, at both the top and bottom of

their respective gamuts.

Accomodation of the Harpsichord Range: Key Transposition

All of the harpsichord concertos with extant models are set a tone below their original

key, and similar transposition appears to have been involved in some of the other

 

’These extant autograph parts are listed in NBA KB IX/2, 209-210.

65

concertos as well. Much has been written about probable reasons for these key
transpositions,ts but most likely transposition was necessitated by the differences in range
between the violin and the harpsichord. For the solo violin, the range normally reaches
to e’”, and only rarely is exceeded. The available range of the cembalo, however, was
more limited. In some Bach works, the keyboard discant rises only to c”’, while in
others to d’ ” . This range difference has been used by some to determine the authenticity
and dating of these works.7 However, the extension of the harpsichord’s range to d”’
occurred well before the time of composition of these concertos: d’” is found in the
Partitas (Clavieriibung I) from the late 1720’s, while in the Sonatas for Violin and
Clavier, BWV 1014-1019, Sonatas l, 2, 4, and 6 all rise to d”’ and Sonata 3 reaches
to d"”. These violin sonatas had already appeared by 1725, and are often accredited
even earlier to Kdthen. Hence the enlarged harpsichord range to d”’ may be affirmed
by the early Leipzig years at the latest. For all of the present concertos, d”’ is not
exceeded, and accordingly may be assumed to be the upper limit available to Bach at the

time.8 It thus seems reasonable to assume that the downward transposition was born out

of the necessity to accommodate this range.

 

6A fairly thorough discussion of transposition and range in Bach’s concerto
transcriptions may be found in Siegele, Kompositionsweise, 105-111, 118-123. Most of
the discussion centers on consideration of the upper limits of the keyboard gamut.

7Paul Waldersee was one of the first to suggest that the upper range of the keyboard
might serve as a clue to dating. Waldersee, BG 31, x.

8In The Art of Fugue from the late 1740’s, e”’ appears in the allio modo fugue for
two claviers, which may suggest a further widening of the available range. That this
higher pitch never occurs in these concertos attests to their completion by 1741.

66

It nonetheless remains true that many range difficulties could have been
accommodated by retaining the original key and altering those passages where the
available range is exceeded. Indeed some passages required such alterations despite the
lowered key. In the second movement of 1041/1058, for example, the solo violin rises
through a broken arpeggio to g’”, well above the usual upper range, and above the
upper limit of the harpsichord even when transposed down a tone. By inverting the
figure in the cembalo discant, Bach accomodates the passage to the available range, but
retains some sense of the original upward motion by adding part of the ascending figure

into the LH (Ex. 2-3).

Ex. 2-3. 1041/1058, II, mm. 14-15

1041
violin

 

 

 

 

all

 

 

105 8
cembalo

 

Most commonly, problems of range are accommodated simply by octave shifts up or

down, as in Ex. 2-4 where the solo violin rises to f’”.

67

Ex. 2-4. 1041/1058, III, III. 80

1041
violin

 

1058 7 --==E=—5r__EEE=E—=E_=="='=__-=_-.:_-

cembalo

 

or in the first movement of the same concertos where the bass in the model descends to

C (Ex. 2-5).

Ex. 2-5. 1041/1058, 1, mm. 130-133

  

y,__———_——__—_ _—v—- ——_—_
.1 .
—_——- -—t__———__-— n—-—————

..———-——_——_ —
' _—- ——-_—_—-—4

1058
cont.
cembalo

It should be noted here that the octave transposition occurs earlier than actually required,

so as to preserve smoother voice leading into the B" (which, if untransposed, would be

68

beyond the compass of many harpsichords). Further, it is clear that the left hand does
not merely follow this altered bass, but, in mm. 130-132, softens the leap of a 7th
(present in the continuo) with the paired eighth notes. This is but one illustration among
many of Bach’s careful attention to melodic motion and voice leading. In other words,
the process of transcription, even when it involves something as routine as octave
transfers to keep within the range, is never done mechanically.

Generally, the solo violin in the violin concertos is distinguished from the first
ripieno violins by two means: the greater brilliance of the violin writing, especially in
solos, and by reserving the highest notes of the violin range for the solo violin. Since
the violino principale is transferred to the harpsichord discant, it is this part which most
frequently encounters problems with the upper extremity of the range and necessitates
either octave transposition or a lowered key. Fortunately there are many instances where
lowering the key one tone is sufficient by itself to solve range difficulties without
violating the integrity of the original melodic line. Such occurrences seem to strengthen
the view that downward transposition is almost surely undertaken principally to satisfy
range requirements. For example, in the first movement of 1042/1054, the solo violin

9,,

ascends to e just before the da capo return of the opening tutti. Transposing the key
down a step allows the passage to be preserved unaltered. The musical effect would be
altogether different if the original key had been retained and the passage rendered an

octave lower (Ex.2-6).

69

Ex. 2-6. 1042/1054, 1, mm. 120-122

udagio

1042

 

1054 Cembalo
original key,
lowered octave

\

   

 

 

V J

A similar instance occurs in 1041/1058 near the end of the final tutti in the first
movement, where a two-octave ascending scale in the solo violin culminates on e’”, a
pitch beyond the upper limits of the harpsichord. In the transcription, the lowered key

9,,

permits the passage to ascend to its climax on d , which is a better solution than

retaining the original key of A minor and adjusting the scale passage down an octave

(Ex. 2-7).

70

Ex. 2-7. 1041/1058, 1, mm. 164—166

1041

1058
Cembalo

 

This example is also interesting for other reasons, however. The lowered key which
preserves the solo melody in the discant also moves the bass (C in 1041) below Bach’s
customary range for the harpsichord, thus necessitating upward octave transposition of
the bass in mm. 165-166. In essence, the solo line in the discant is given priority and
is preserved at the expense of the bass. Of greater significance, however, is the fact that
the relationship between solo and ripieno strings has also changed here. In 1041, the
first violins double the violino principale until the last three sixteenths before the peak,
when the ripieno violins drop a step and continue in thirds with the solo violin. In effect
this ensures that the solo violin remains preeminent musically, attaining the highest point
in the tessitura so as to stand out. This relationship is significantly altered, however,
when the solo violin is replaced by the harpsichord, an instrument of different timbre,

in 1058. No longer in danger of being lost amid the first violins, the scalar ascent to

71

d’” is now doubled fully by first violins, with the drive to the downbeat of m. 166
intensified by the simultaneous addition of the descending scale in the left hand.
There are far more examples of Bach’s careful attention to range problems than
need be cited here. While some instances present no difficulties, others show deliberate
concern to present melodic material in the best possible manner. Ina few cases, this
may actually mean raising melodic material an octave to give it greater prominence. In
the first movement of 1049, mm. 217ff., the concertante violin takes over the, theme
from the two recorders, with minimal ripieno accompaniment. In the harpsichord
transcription (1057), the violin passage is given to the cembalo discant in the customary
manner, but raised an octave, thus reaching to the highest pitch of the harpsichord (Ex.
2-8).° While this gives the theme greater prominence, it also lends greater brilliance

to the harpsichord.

Ex. 2-8. 1049/1057, 1, mm. 217-221

1049
violin
recorders

1057
Cembalo

 

 

’That this upward transposition results in the appearance repeatedly of d”’, coupled
with the general avoidance of e’”, seems to confirm that d’” was indeed the highest
pitch available on Bach’s harpsichord.

72

A parallel example may be found in the third movement of these same two concertos,
at mm. 183-188. Particularly in this instance, raising the cembalo discant an octave
gives the harpsichord a distinct timbral advantage. Whereas in 1049 the solo violin plays
below the second recorder, in 1057 the same musical material is elevated an octave in
the cembalo discant and thus gains audibly by now sounding above the recorder, which

remains in the same octave as in the model (Ex. 2-9).

Ex. 2-9. 1049/1057, III, mm. 183-188

1049
violin
recorders

 

1057
recorders
cembalo

 

 

While adjustments to accomodate the instrumental range are most common in the
harpsichord discant, modifications are occasionally required in the bass as well. The
range limit of the bass as determined from these works is somewhat unclear and
puzzling, however. It might be presumed that the lowest pitch available on Bach’s

harpsichord was C, since in every case where the model concerto employs low C the

73

transcription transposes the note or surrounding passage up at least an octave, thus
avoiding the BB" which would result when the key is transposed down a tone.‘0 By
thus consistently and deliberately avoiding BB or BBb, the implication follows that C was
the lowest available pitch, which is consistent with the gamut of many harpsichords of
the day. However, in several of the harpsichord concertos, pitches below C do appear,
reaching even as low as GG. Of the seven solo concertos in the autograph manuscript
P234, the bass dips below C in five of them (1052, 1053, 1054, 1055, and 1056). If the
keyboard on Bach’s harpsichord did in actuality reach below C, then why the octave
displacements whenever the model used C?

During the first half of the eighteenth century, the range of the harpsichord did
gradually increase, a fact which has led some scholars to suggest that certain keyboard
works may be datable according to the range employed.“ While this presumption is
attractive, it is also problematic, inasmuch as the lack of a certain pitch in a piece does
not substantiate its absence from the keyboard.12 The question thus remains of just
what notes were present in the bass of Bach’s harpsichord. Examination of the bass

range in some of Bach’s solo harpsichord works may shed some light on this question.

 

10Instances of C or C” in the bass may be found in the violin concertos as follows:
In 1041, I: mm. 20, 55, 64-65, 80, I39, 166; 11: mm. 34, 46; in 1042, 1: mm. 70, 91;
II: 35; in 1043, 1: mm. 62, 69, 87; II: mm. 3, 6, 15, 43, 46, 49; III: mm. 40, 59, 83;
in 1049, 1: mm. 203, 217, 221, 235, 237, 239, 243; 11: m. 31; 111: mm. 110, 144. In
every case without exception the pitch is raised in the transcription, thus avoiding BB or
BBb when transposed.

“Generally, the increase of the upper range of the harpsichord has been the focus
of this consideration. See BG 31, x.

12For example, if a work does not contain the pitch d”’ or AA, one cannot assume
that this pitch was necessarily absent on the intended instrument; it is equally possible
that the composer deliberately chose not to use such pitches, even if available.

74

The solo harpsichord part of Brandenburg Concerto V, written no later than
1721, never crosses below C, and the same is true of all 24 Preludes and Fugues that
comprise the first book of the Well-Tempered Clavier of 1722. The Violin Sonatas with
Obbligato Cembalo (BWV 1014-1019) have been dated by some scholars to no later than
c. 1723, and in most of these the lowest pitch is C or C”. However, in the first and
third of these sonatas (1014 and 1016 respectively), BB is encountered a total of six
times.13 The six keyboard Partitas appeared between 1726 and 1730, and are most
interesting in regard to the range issue. All of them have a bass range which exceeds
C: #2 and #3 each reach AA; #6, BB; and Partitas 1, 4, and 5 each descend all the way
to GG. Of further note, however, is the observation that Partitas #3 and #6 also exist
in an earlier form in the Clavierbiichlein for Anna Magdalena Bach of 1725. In this
slightly earlier reading, C is the lowest pitch, though it is clear from the following
comparison (Ex. 2-10) that BB is the desired pitch in at least one passage:

Ex. 2-10. Clavierbiichlein for A.M. Bach - Courante, mm. 11-13

—.¢—.-.

. .....__'_...

A ' ’

 

a. .o—a——— ..-. - .

Partita #3 in A Miner - Courante, mm. 11-13

 

 

13The fact that BB occurs only in these two sonatas may call into question the dating
of these works to 1723, especially in light of the fact that the only surviving source
materials for these sonatas are from the Leipzig years. It may be worth noting also that
1014 displays a musical and structural character which is more advanced and a cembalo
part which is more independent than that found in many of the other violin sonatas. This
fact, coupled with the broader range, may suggest a slightly later date.

75
The Well-Tempered Clavier, Book II, dates from 1738-42 and does not exceed C through

the first eighteen Preludes and Fugues. However, in Fugues #19 and 20 the bass drops
to AA, and in Prelude #23 and Fugue #24 employs BB. Lastly, while most of the
Goldberg Variations, BWV 988, from 1741-2 remain above C, instances of AA occur
in Variations 14 and 18, and GG appears once at the final cadence of Variation 24.

The foregoing would seem to suggest that the range of Bach’s harpsichord did,
in fact, widen during his career, particularly during the Leipzig years. The extent to
which this fact may be used to date the keyboard works remains open to debate. But
what appears certain is that Bach had indeed written works for harpsichord with a
compass reaching to GG even before the time of the harpsichord concertos. The
question of why BB or BBb was avoided when transposing the key downward thus seems
all the more clouded.

It might be noted that while only AA, BB, and GG are used in the solo keyboard
works cited above, in the concertos these same notes plus BBb and GG# appear, i.e. all
the chromatic pitches from GG to C. In most if not all of these instances in the
concertos, the notes below C appear to have been added in at the time of the writing of
the autograph. This can be verified by comparison of parallel passages in the concerto
versions with either the model, when extant, or with other parallel versions, such as
Cantata movements. Often the pitches below C appear as the result of interpolated
embellishment, octave doubling, or filler via altered rhythmic treatment. In some cases,
it is conceivable that performance on a harpsichord without these low keys could be
realized by playing the pitches an octave higher or leaving out a doubled low note, such

as in Ex. 2-11.

76
Ex. 2-11. 1054, 1, mm. 73-77, cembalo

 

 

While it remains preferable to play the GG in m. 76, and thus preserve the parallelism
with mm. 70—71 and 73-74 (octave leap up to the sixteenths), it is also possible to omit
the G6 without seriously undermining the musical integrity of the passage. In the
parallel measures in 1042, the model for 1054, the continuo bass after which the left
hand of 1054 is fashioned appears an octave higher. It is evident in the autograph of
1054 that Bach first began to sketch in the continuo at this higher octave, but then
crossed it out and inverted the octave to a descending rather than ascending direction,
thus positioning the subsequent sixteenths in the lower octave in both the continuo and
the cembalo bass. Low GG was then also added, tell-tale signs of which are the overlap

of note stems with the first violin part on the next lower staff (Ex. 2-12):

77
Ex. 2-12. 1041, I, 76-77

--- mu —— —— — ——— ——_—‘——_-
r.-.-- ——— -—— —- —_—-_’-_—» - __—-———u
u .-_———-——— -— —-——--————— ——————--
wv_———_———————-—-r—---———--———-

 

   

a

7 2 215155:

The low GG thus allows the ascending octave leap to be retained intact in the left hand,
with the continuo doubling at the same octave in the following measure. Musically this
is satisfying; practically, it is possible only on a harpsichord with a range extending at
least to GG. A further passage may be shown which requires an instrument with low

GG but does not lend itself to any other performance accomodation (Ex. 2-13). In this

78

case, a scalar descent to GG admits to no other musically satisfying option, and clearly

requires BB”, AA, and GG.“

Ex. 2-13. 1052, 111, mm. 271-272, Cembalo

 

During the first half of the eighteenth century, harpsichords in Germany could
be found with a variety of ranges. Harpsichords made in the seventeenth century such
as those by Ruckers typically had a range of C to c’”, and their highly respected fine
tone quality kept them in use throughout the eighteenth century." In the early
eighteenth century, the range of the harpsichord in Germany was expanded to five
octaves, typically FF to f’ ”, though considerable variety was still to be found. Frank
Hubbard, in his monumental study Three Centuries of Harpsichord Making, cites
dispositions of selected 18th century harpsichords, and shows that even instruments built
between 1710 and 1737 offer diverse ranges including FF-f’”, GG-c”’, GG-d’”, and
C-d’”. Hubbard states that FF-f’ ” is the usual range of German harpsichords, but more

limited ranges may occur occasionally on even the largest harpsichords. Eighteenth-

 

“This passage does not appear at all in 1052a, the parallel version of this concerto
possibly by CPE Bach. The corresponding passage in 1052a is an empty measure
marked ”ad libitum.” See BG xvi, 313.

1’By way of example, one such large Ruckers instrument made in 1651 is mentioned
in Handel’s will. Hanns Neupert, Harpsichord Manual (Kassel: Barenreiter, 1968), 38-
39.

79

79’

century documents cited by Hubbard mention instruments with a gamut of C-c , while
another exemplar by the builder J. A. Hasch from 1723 (now in the Musikhistorisk
Museum, Copenhagen) which may be close in disposition to a harpsichord thought to be
close to Bach has the range FF-c”’.16 Thus considerable variety appears to be common
among keyboard instruments of the time.

In the light of these qualifications, the question of the bass range of Bach’s
harpsichord returns. Two seemingly contradictory observations have surfaced: first, that
Bach avoids notes below C when transposing a concerto down a tone; and second, that
in the autograph P234, pitches to GG occasionally appear which apparently were
incorporated into the transcriptions during the writing of the manuscript. Table 3-1
below lists the appearances of all pitches below C, or the lowest pitch if C or above, in
each movement of the concertos in P234, plus the three concertos for two harpsichords

(1060-1062). The double concertos are listed first, since they appeared first, followed

by the works with extant models, and lastly the remaining concertos without models.

 

16Frank Hubbard, Three Centuries of Harpsichord Making (Cambridge: Harvard
University Press, 1965), 178-180, 185, 329, 331.

80

Table 2-1: Bottom pitches in harpsichord concertos

VVork

 

1060

1061

1062

1054

1057

1058

105217

1053

1055

1056

hdodel

 

1043

1042

1049

1041

hdov.

I
H
HI

H
[H

H
[H

U
IH

H
[U

H
[H

H
HI

H
HI

H
IH

HI

 

17Additional occurrences of BB”, AA, and GG in 1052 appear in the BG edition
(mm. 22, 134-136, 138) which are not in P234. The BG was prepared from a later parts
copy by two unknown copyists (Mus. ms. St 125), which, though dependent on the
autograph, departs from it at a number of points. The pitches recorded in the table

Puch

DUO GOO

000

CH}
cl
cl

GOO

COD

lkA
138
BE?
BB

I38
(“3'

bdeasure

76

117
83,271,272
103

272

72

44.86.302.343
226

17.44
30

9.11.13
14

reflect the reading in P234. NBA KB VII/7, p. 41.

81

The table makes a couple of observations very apparent: first, that none of the
double concertos, which come from the earlier years of Bach’s association with the
Collegium, reach below C, even though 1043, the model for 1062, uses C or C” a dozen
times. In every instance, the passage containing C is raised an octave to avoid BBb when
transposed. Second, that among the solo concertos with extant models, only once does
a pitch below C appear. That sole example, in 1054, is the low GG cited in Ex. 3-9 and
which is doubled by the G an octave higher, easily omitted if necessary. The examples
of pitches below C thus are basically confined to the four concertos without models. Of
these, three concertos include movements which also appear in parallel versions: 1052
I and II in Cantata 146, III in Cantata 188; 1053 I and II in Cantata 169. III in Cantata
49; and 1056 II in Cantata 156. However, in none of these cases is downward
transposition involved; either the keys are the same in the two versions, or the concerto
key is actually higher than the cantata version.18 Thus the seemingly contradictory
conditions of avoiding BBb while also using pitches to GG do not occur in the same
composition (with the one exception of the GG in 1054). This strongly suggests that the
transcriptions may well have been written with two different harpsichords intended, one
reaching to C, the other to GG. It would be an attractive conceit to suggest that those
concertos without pitches below C predated the concertos with a lower range, but given

the layout of P234 and the fact that 1057 and 1058 are the last two concertos in P234,

 

l"These parallel versions are themselves likely derivatives from still earlier, now lost,
concerto models. There remains then the possibility that the lost model was set in some
other key, and both the cantata and concerto versions were transposed. Regrettably,
there is virtually no way to ascertain the original key if different from what has survived.

82

this presumption is not tenable.19 It might be better. to suggest that perhaps when Bach
transcribed 1057 and 1058, he had in mind other possible performance venues in addition
to the Collegium, venues where a harpsichord of smaller compass would be available,
and accordingly used a gamut which was possible for the Collegium and elsewhere.
Definitive conclusions to such questions cannot be realized, but the observable
data nonetheless invites speculation. Of the harpsichord dispositions cited by Hubbard,
those which most closely match the works in question are C-d’” for 1054, 1057, 1058,
and the three double concertos 1060, 1061, and 1062; and GG-d”’ for 1052, 1053,
1055. and 1056. The range of the latter suggests a harpsichord with a ”short octave”
of the type in which the lowest key looks like BB but was tuned to sound GG. The
black keys C‘l and D” were tuned to AA and BB respectively, unless the instrument was
fitted with split keys for these two notes, in which case the front half was AA and BB,

the back half 0' and E" (Fig. 2-1).

 

 

 

 

Fig. 2-1. Illustration of short octave

Even if equipped with split keys, both AA" and BBb remain absent, pitches which are

required in 1052, 1053, and 1056. However, since both GG and GG" do not appear in

 

19It will be recalled that the end and beginning of adjacent concertos in P 234 share
the same fascicle of the manuscript; hence a concerto could not have been transcribed
earlier and added into the manuscript at a later time.

83

the same concerto, nor similarly BB and BB”, it is possible that the pitches not available
via short octave could be realized by scordatura, i.e. tuning the GO to GG” or the BB '
to BB". Given the absence of FF or pitches above d”’, this seems more probable than
assuming a full five-octave keyboard FF-f’ ”.

It may be remembered that in June of 1733, announcement was made of the
acquisition of a new harpsichord for the Collegium, "the likes of which all here have not
yet heard.” Could it be that the special character of this instrument was in part due to
a broader compass, reflected in these concertos?

Finally, one last observation concerning the cembalo bass range may be in order.
While the relationship of the cembalo bass and continuo will be examined thoroughly
later, it is worth noting here that in those instances where the left hand part doubles or
parallels the continuo and the left hand drops below C, the continuo bass line does not
descend below C. Hence while the cembalo range extends to GG. the continuo range

never crosses below C. Ex. 2-14 shows a very clear illustration of this phenomenon.

Ex. 2-14. 1052 1. mm. 83-84

 

84

This continuo part would have been played by a lower string instrument, most likely
cello or violone. In either case, the lowest notated pitch on these instruments was C,
and so the continuo line is limited to C or above.” The fact that Bach carefully
observed this boundary for the continuo, while exceeding it for the cembalo.
corroborates that the use of GG to BB was not accidental or haphazard, but deliberate,
reflecting a particular keyboard gamut. As noted earlier, one of the unanswered
questions in these concertos is whether a second harpsichord was employed for the
continuo part. The only real evidence supporting such a hypothesis is the existence of
a figured bass part for 1055, which by itself does not prove the practice. However, if
a second harpsichord were used to provide a bonafide basso continuo, then this
instrument may well have been of smaller compass than the solo instrument, i.e. C—c”’
or C-d’”.

Occasionally, key transposition also necessitates alteration of ripieno parts, as
when lowering the key forces a passage beyond an instrument’s compass. A particularly
interesting example of this is found in the first movement of 1043/1062, where the
ripieno violin parts in 1062 are raised an octave in order to avoid the unplayable low f
(Ex. 2-15). As a by-product of the transposition, m. 70 is also restored to its intended

reading. In 1043, the low f in m. 70 could be played only by violas.

 

”Laurence Dreyfus distinguishes three different string bass instruments, plus the
Violoncello, which served Bach as continuo instruments in Leipzig: the Violono grosso,
a 16-foot instrument sounding an octave below its notated pitch. a violone with a low D
string, and a violone in G. None of these extend below C, except the Violono grosso
which is notated only to C though sounding an octave lower. Laurence Dreyfus, Bach ’s
Continua Group (Cambridge: Harvard Univ. Press, 1987), 132-151.

85
Ex. 2-15. 1043/1062, I, mm. 69-70, ripieno parts

1043

1062

 

Numerous similar examples of octave shifts to accommodate the instrumental ranges may
be found in these concertos, affecting every instrument to some degree. More
importantly, many of these alterations support the hypothesis that transposition was used
to accommodate the lower top range of the harpsichord. It can also be demonstrated that
this accomodation of range, when necessary, was not undertaken haphazardly, but with
care for the melodic motion and the technical limitations of the instruments involved.
Bach’s thorough knowledge of the instruments for which he wrote is reflected in the
avoidance of unduly awkward or unplayable passages. This might mean circumventing

an instrument’s ”poor? notes, such as f‘" on the recorder (1057, I, m. 186), or simply

86

adapting a passage to the technical capabilities of an instrument. A most enlightening
passage which well illustrates the latter occurs in the fourth Brandenburg Concerto
(BWV 1049). at mm. 98-102 of the first movement. Here the solo violin’s arpeggiated
figuration is altered when compared with the parallel passages at mm. 120-124 and mm.

258-261 (Ex. 2-16).

Ex. 2-16. 1049, 1, 98-102 and 120-124, solo violin

98—102

120-124

 

Clearly the passage at mm. 98-102 ought to read as follows (Ex. 2-17):

Ex. 2-17. 1049, 1, mm. 98-102, reconstructed

 

It is notable that this is in fact the way the passage appears in the transcription, 1057
(Ex. 2-18), i.e. the ”correct" version appears in 1057, the ”altered” version in the earlier

model.

87

Ex. 2-18. 1057, 1, mm. 98-102

 

Siegele suggests that the reason for the alteration at m. 102 in BWV 1049 is that the
unaltered figure cannot be played on the upper three strings, as is true for the
surrounding measures;21 also the violin figuration in these measures uses first position,
and c" to a’ cannot be played in first position. None of these restrictions apply to the
harpsichord, of course, and so the ”normal” reading is restored, as it were, in 1057.
Considering that the fourth Brandenburg was written more than fifteen years before the
harpsichord concerto, it is all the more remarkable that the transcription "corrects” this
passage to its original conception, thus aligning it with parallel passages. This
remarkable example also attests further to Bach’s attention to detail; the process of

transcription was in no sense mechanical or thoughtless.
Idiomatic Adaptation to the Cembalo

Although the musical material for much of each transcription was taken over
directly from a source model, the part of the harpsichord often required considerable
adaptation, in large part due to the different playing technique required and the
difference in the sound medium itself. While the violin intrinsically has the capacity to

sustain long

 

2‘ Siegele, Kompositionsweise, 125, fn. 37.

88

notes and fashion a broad legato, the harpsichord is limited by its characteristically brittle

attack and rapid sound decay, and this timbral difference had to be accomodated.
Recognition of the differences in technical and timbral character between the

violin and harpsichord is reflected already in changes in Bach’s tempo markings, as

illustrated by a comparison of tempi in the concertos and their models:

Model Transcription
1041 (Allegro) 1058 (Allegro)
Andante Andante

Allegro assai Allegro assai

1042 Allegro 1054 (Allegro)
Adagio Adagio e piano sempre

Allegro assai Allegro

1049 Allegro 1057 (Allegro)
Andante Andante
Presto Allegro assai

1043 Vivace 1062 (Allegro)

Largo ma non tanto Andante e piano

Allegro Allegro assai

While the same tempo markings prevail in most parallel movement pairs, several
significant changes are evident in 1057 and especially 1062. The opening movement of
the latter, marked ”Vivace" in alla breve meter in 1043, is changed in 1062 to common
time without tempo marking, suggesting a slowing of pace to "tempo ordinario"; in the
middle movement, "largo, ma non tanto" yields to "andante e piano"; and in the final
movement ”allegro" becomes ”allegro assai”. These changes in tempi and expression
suggest accomodation to performance parameters better suited to the technical limitations

and starker tonal idiom of the harpsichord. For the second movement particularly, the

89

slightly brisker pace of Andante is best understood as tacit recognition of the inferiority
of the cembalo to sustain tone when compared to the violin.

Beyond tempo markings, one finds much more attention given to performance
directions in these works than is typical: dynamics, phrasing, staccati, trill signs, and
a more detailed use of notation are all in evidence. In some cases, these represent
concessions to the technical limits of the harpsichord. Long sustained notes are well
suited to the violin, but fade quickly into silence on the cembalo. They are accomodated

to a keyboard instrument by energizing the tone with a trill (Ex. 2-19).

Ex. 2-19. 1049/1057 I, 24-29

1049 violin

 

1057 cembalo

 

Dynamic markings appearing in the models usually are duplicated quite faithfully in the
transcriptions, but there are instances where modifications or additions do appear, as for
example in the double concerto 1062. Here in the third movement are a number of
dynamic markings, both piano and forte, which are absent from the model (1043). In
some instances these do little more than reinforce the return of the tutti, or soften the

accompaniment in solo episodes. Two examples particularly stand out, however: the

90

forte at m. 72 to underscore a unison cadence in all parts; and the use of not just piano
but pianissimo in m. 22 of the first movement at the first solo entrance. Occurrences
of p and f are infrequent enough in this music, but pianissimo is particularly rare.
Bach’s direction pp at this first solo entrance (rather than the p of the model) is no doubt
intended to insure that the two cembalos project adequately over the ripieno. Less
frequent are instances where dynamics found in the model are not transferred into the
transcription. Most commonly, these are dynamics which first appeared in the solo
violin part which, when transferred to harpsichord. become irrelevant. Examples may
be cited from the third movement of 1042, where dynamic changes marked in the solo
violin part are absent at the corresponding passage in 1054 (Ex. 2-20). Since the
harpsichord is capable of little real dynamic variation. a limitation which is often
ameliorated by other means such as agogic accents. such dynamics as do appear in the

transcriptions are applied to the ripieno string parts.

Ex. 2-20. 1042 III, 62-65. violin

 

 

 

In general, Bach’s score gives detailed attention to exact expression marks,
including slurs, trills. staccatos, and frequent dynamic indications. While many of these
notational details are copied unchanged from the violin concerto model (1042), there are
instances where the reading in the transcription is altered toward greater notational

precision. a tendency which is observable in Bach’s later years. Thus the opening triadic

91

theme of the first movement is notated as quarter notes with staccato dots in the violin

concerto version but as eighth notes-eighth rests in the harpsichord reading (Ex. 2-21).

Ex. 2-21. 1042/10541, mm. 1—2

   

1042 Viollno mama :2"- ,Ed- -.———_-;5__—— — ——————

Violine l
Violine I] ’j‘ ; . .
Viola

Continua

1054 \‘iolino r.
Viniino ll.
"ioltl.

Coulinuo.

Cembalo.

 

In the second movement of the same work, the even eighth notes of mm. 23-24 and 48-
49 in the violin concerto are replaced by syncopated Iombard rhythms in the harpsichord

version (Ex. 2-22).

92
Ex. 2-22. 1042/1054 11, mm. 23-24

1042, Violin 221:; =:;==:‘:r=:a='1‘"_=- ==E:._.-. 3....--

1054, cembalo

 

It has been suggested that perhaps Bach intended the passages in 1042 (mm. 23-24, 38-
49) to be performed in the same way, i.e., with Iombard rhythm as in the harpsichord
concerto.” However, since there are other passages in the same movement where '
sixteenth notes appear in slurred pairs in 1042 but are not in Iombard rhythms in 1054,
this suggestion seems unlikely; the reading in sixteenths either is preserved (mm. 14 19),
or is embellished with 32nds (m. 35). In essence, the Iombard rhythm becomes but one
type of embellishment of the sixteenth note pattern, one which perhaps takes greater
advantage of the harpsichord’s intrinsic ability to declaim dotted rhythms with special
clarity.

Much of Bach’s purpose in adapting these transcriptions is to make the writing
more idiomatic to the keyboard, particularly in solo passages. For the discant, achieving
this goal is largely accomplished via two means: first, the adaptation of violinistic
figuration to the harpsichord, and, second, the embellishment and filling in of melodic
outlines into more brilliant passagework. Adaptation of idiomatic string writing,

including rapid repeated notes, bariolage, wide-spaced arpeggiations, and double and

”Schulze, Forward to Konzert D-Dur. n.p.

93

triple stops, is encountered frequently. As but one example, in the opening tutti of the
E Major violin concerto (1042), the solo violin doubled by ripieno strings features
several passages of rapidly repeated notes, a device well suited to the violin but
uninviting on the keyboard (Ex. 2—23). The passage is accomodated to the harpsichord
simply by substituting broken octaves somewhat in the manner of a tremolo, a common
keyboard idiom. Since the first violins of the ripieno doubled the solo violin part in the
violin concerto (1042), and hence are transferred unchanged into the cembalo concerto

(1054), the two readings may be easily compared in 1054.

Ex. 2-23. 1054, 1, mm. 4—6, 9-11

   
     

. ____.- ....u..v..
: __ ___.___=_---_-

__.__— __..'
- >—- 2 . . - »'—"

.—-.---_-— —
WW—

—7 — —— —
‘1

 

 

 

94

It is worth noting here that not only do these ”tremolo" octaves find application in the
discant, but also in the left hand part of mm. 9 and 10. Although the harmonic content
remains unchanged, the result is better use of keyboard technique and a balancing of the
broken octaves of the discant in contrary motion. The second half of m. 11 also
provides a good example of embellishing a cadence with a brilliant two-octave scalar
flourish which not only suits the keyboard well, but also very effectively introduces the
first solo episode, which immediately follows.

Passages which are most idiomatic to the violin occur, naturally enough, in solo
episodes rather than in tuttis. One of the most interesting examples of Bach’s keyboard
adaptation occurs in an extended solo passage in the first movement of 1042/1054, mm.
57-70. Here the original solo part is comprised of 13 bars of broken intervals in
sixteenth notes, a figuration which is very idiomatic to the violin but is not particularly
effective on the harpsichord. Two versions of this passage may be discerned in the P
234 autograph, representing two stages or layers of composition. Bach first wrote out
a right hand part which was close to the original violin part but with some concession
to keyboard technique, but later replaced this with a more idiomatic reading, recorded

in tablature below the bass system (Ex. 2-24).23

23The first version, which appears in regular notation in P 234, is the "altere
Lesarte" published in the appendix to BG XVII. 316, and is reproduced for the following
example.

95
Ex. 2-24. 1042/1054, 1, 57 ff.

1042. solo violin (BG)

.-

. -- ~ - - _ -‘-‘—- '- m—v —

-—‘ =—--=-=-‘=-‘=-=---- _J—A-J-J uu-‘= -- '- l- -- .

lI—--&—=—-——=—-— ———=:====—-—---—-—---—-- —_, --= -—=_--------
' -- - - ,

_n—. .‘_-—a—‘ a.—;—.—— r. n. -— -— '1 -— -— -—— '1 -- -- nun-—
.-_————-q—--_=-P———_-‘-—_———=‘--————-—a—---——--
_-_ — ——---————— ——————-——-::—--— -==E

‘——=-'=_'l— _==.=—=—-==

—===-_===_§g== ' = —

 

’- ‘/———--‘.——:—=: _——_ --

~Jv-___ —— ..-—— _V _—
,_____._'__:,:_— — —=—=—=—_——-—

U." — __ ——_——_—-———-____-

When compared to the very minor changes made in ripieno parts, the alterations found

here in the cembalo discant are quite striking.24 The presence of two distinct layers of

 

24Pook has suggested that this passage is not really an adaptation of violin figuration
to keyboard, but rather is Bach’s "final intentions as to the manner of embellishing
certain notes and passages in performance," i.e., it represents a written-out realization
of what was expected to be played from the original, whether on harpsichord or violin.
This viewpoint, while interesting, is not very convincing. Even Pook admits that such
a written out performance realization would be exceptional if not unique in Bach.
Wilfrid Pook, "Bach’s E Major Violin Concerto Reconsidered," Music and Letters
Xxxvii (1957): 53, 57—58.

96

reworking the original material demonstrates Bach’s determination to give the
harpsichord the individuality of a solo instrument. In the first attempt, the process of
transcription is more mechanical, resulting in a figuration which still betrays its violin
roots. The subsequent reworking is more radical, and, though still tied to the same
harmonic background, moves the harpsichord further from its model. How much time
elapsed between the two versions is unknown; perhaps the second version owes its
provenance to the composer’s unsatisfactory judgement of a rehearsal or performance of
the first version with the collegium.

A somewhat similar solo passage involving two-stage alteration of the original
violin figuration appears in the same movement at mm. 82-92 (Ex. 2-25). Again the
first stage displays a very close adherence to the original string figures. characterized by
repetitive 2nds and 3rds which are retained for the most part in the transcription. The
second stage, again recorded in tablature, widens the repetitive string figures, resulting

in greater melodic interest as well as filling in the harmony.

97

Ex. 2-25. 1042/1054, I, 82-86

1042 (BG)

 

 

 

 

1054:
lst version”

 

2nd version

 

At m. 88, the change of figuration is even more significant. In the violin model, the

repeated arpeggio patterns widen to 11ths and 12015, awkward on a keyboard in this

 

2"’From ”altere Lesarten" in BG, XVII, 316-317.

98

brisk tempo. Bach’s first thought on the matter reduces this wide compass to 6ths,
which fit comfortably under the player’s hand; but in the reworked final reading, the
wider compass is partially restored to octaves and 7ths, which are still negotiable but
more demanding for the player nonetheless. The remaining melodic figure in each half

bar is also made more interesting and less mechanical (Ex. 2-26).

Ex. 2-26. 1042/1054, I, 88-91

1042

 

 

1054:
lst version

 

2nd version

 

99

Arpeggiated figures of wide compass, so characteristic of violin writing, appear
frequently in these works, and are accomodated to the harpsichord in a variety of ways.
At mm. 23-24 of the concerto movement cited above, one may gain some sense of how
wide a span on the keyboard Bach felt was reasonable in brisk tempo. The original
violin figure begins with an arpeggio spanning a 6th but widening to a 12th; the altered
passage in 1054 begins in the same way, but restricts the widening compass to no more

than an octave (Ex. 2-27).

Ex. 2-27. 1042/1054, 1, 23-24

1042

1054'

 

That the octave represents a reasonable hand span for Bach is further illustrated in a
somewhat similar example from 1041/1058 III, where the figuration of the solo violin
in the model spans a 12th but is reworked and reduced to no more than an octave in
1058 (Ex. 2-28). Also noteworthy is the left hand part, which is embellished with

arpeggiation but restricted to an octave’s compass.

100
Ex. 2-28. 1041/1058 111. mm. 82-85

1041 v1 and cont.

    

 

1058 cembalo

 

There are also instances in these transcriptions of wide-spaced arpeggios where
the original compass is retained but positioned in such a way that the passage is divided
between the hands, thus facilitating its playing. In the third movement of 1054,
descending leaps of a 12th occur in sixteenth notes between ascending arpeggio figures
(Ex. 2-29). At the marked tempo ”Allegro,” these would be unplayable except for the

assistance of the left hand.

Ex. 2-29. 1042/1054, III, mm. 19-20, 23-24

1042

 

1054

 

101

It is worth noting in mm. 23-24 of the above example that Bach raised the bass part (the
original continuo bass line) an octave in m. 24 only, making it possible for the left hand
to reach the d”’ and thus facilitate playing the passage. This octave shift is notable in
that it illustrates a transposition which has nothing to do with range, but results solely
from the careful consideration of the playing technique of the instrument.”

Some of the most brilliant keyboard writing to appear in these concertos is found
in 1057 I, mm. 83—125, where the harpsichordist plays a cadenza-like section alone while
the ripieno strings drop out. As in the immediately previous examples, the melodic
material of the model again contains wide leaps (10ths and 12ths), the playing of which
is accomodated by dividing the passagework between the hands. This division of labor
also facilitates the transformation of sixteenth note arpeggios into more brilliant scalar
runs in 32nds. It is notable that in the autograph Bach is uniformly consistent in the way

in which the passagework is beamed, and that this beaming suggests two-part voice

leading (Ex. 2-30).”

 

26A parallel passage with similar wide arpeggios spanning up to a 12th occurs in the
same movement at mm. 57-60; in each instance the left hand is free to play the lowest
note (or two) of the arpeggio, facilitating its performance.

2"Beams above the notes indicate RH, beams below LH. In the model (1049), the
solo violin part is written with one double-beam per bar and the first three sixteenths
slurred if an ascending arpeggio. There is no correspondence between this slurring and
the hand—division beaming of 1057; rather, the transcription is set anew.

102

Ex. 2-30. 1057 1. 90-102 cembalo

 

Additional examples of violinistic figuration, including rapidly repeated arpeggios.
double notes, and bariolage, all of which must be accomodated to the cembalo, may be
found in every one of these concertos. While the exact manner of adaptation varies in
each case, it nonetheless remains that Bach successfully accomodates the idiom of the
model to the technical resources and advantages of another instrument with integrity and
musical insight borne out of playing experience. To illustrate the latter, it is clear that
in some instances smaller details of notation betraying technical rather than musical
considerations were added later: in 1057 III, a second set of stems and beams was added
to mm. 101—104 to indicate division between the hands for playing ease. It should be
noted here that this division of hand parts has no musical advantage, such as suggesting
two voice parts as in the 1057 example cited above, but it does keep the passage neatly
under five fingers in each hand. The fact that the second beams are added in a lighter

script suggests two possible observations: first, that they were added later, perhaps in

103

a rehearsal with the Collegium; and second, that the manuscript may have been used as

a performance score (Ex. 2-31).”

Ex. 2-31. 1057111, 101-104 P234, cembalo

 

 

The above measures are followed immediately by an extended solo passage incorporating
bariolage. Of all the devices requiring adaptation in transcription, bariolage is the one
most closely aligned with string instruments. As used in the model (1049), each measure
in this passage features a half bar of bariolage alternating with a half bar of broken

intervals for 14 bars (Ex. 2-32).

Ex. 2-32. 1049111. 106-109 violin principale

_ H F555: F -222: H H222: I. gene;

       
   
 

  

 

 

 

In the harpsichord concerto arrangement. this bariolage effect is replaced by broken
thirds in the RH which are better suited to the keyboard. The broken 6ths and 7ths of
the second half of each bar are retained, but altered somewhat from the original version.

The autograph shows two readings for these broken intervals: the first. written in the

 

2”That this added notation is lighter also suggests the possibility of someone other
than Bach writing it in. Obviously, there is no means of ascertaining the source of the
added stems with certainty.

104

upper staff. follows the violin model directly. Evidently Bach found this version
unsatisfactory; he crossed it out and entered a second reading in the lower, otherwise I

vacant staff which achieves greater variety and interest (Ex. 2—33).

Ex. 2-33. 1057 III, 106-110, cembalo

P234

 

first reading:

 

 

second reading:
d

    

One last device which appears relatively infrequently in the model concertos but
nonethless deserves mention is double stops.” Reserved for the solo violin, they lend
further virtuosic character to the string writing. The transference of double stops to the
keyboard is accomplished easily, ’of course, since one of the chief assets of the cembalo

is its ability to play multiple pitches simultaneously. Indeed, in some instances Bach

 

2"In the four model concertos, double stops appear at 1042 I, 95-101 and 111, 82-88;
1043 III, 41-47 and 127-133; and 1049 I, 217-228. Triple stops occur only in 1049 I,
215-216. -

105

simply transfers double stop passages to the clavier directly and without modification
(excepting, of course, the key transposition). Perhaps the clearest example of this simple
procedure is in two passages from the D minor double violin concerto (1043), where
both solo violins play double stops simultaneously for seven bars (III, mm. 41-47 and
127-134). Though this produces a somewhat austere, even bland effect on harpsichords,
Bach makes no effort to embellish or split up the open intervals into more florid figures.
Rather, the double-note intervals of the model are retained in the right hand; as a
corollary, however, this gives more opportunity to develop the left hand to a far greater

extent (Ex. 2-34).

Ex. 2-34. 1062 III, 41-45, cembali

£31111

 

By contrast, in 1042/1054 Bach chose to completely revise and alter a double stops
passage for the transcription, even though the violin model is less static and lends itself
to keyboard performance much more readily than the example cited immediawa above.

In 1042 the double stops alternate with an inverted pedal tone in a broken chordal pattern

106

which for the most part could have been accomodated easily on the cembalo. Instead,
the passage is broken into running triplets, which continue through the solo episode well .

beyond the length of the original double stops (Ex. 2-35).

Ex. 2-35. 1042/1054, 111. 82-89

1042 violin

 

 

 

 

 

 

 

 

 

 

It is perhaps notable that this is the most altered solo episode in the entire movement,
and the only episode in which triplets appear. While the solo string passage is playable
on harpsichord, clearly Bach’s purpose is not merely to reproduce the passage in the
transcription, but to make it more idiomatic and brilliant for the cembalo.

In the first movement of the same concerto, double stops are used to fashion a
short contrapuntal passage implying two independent voices. While performable on

keyboard, the double notes again are broken up into a much more florid line (Ex. 2-36).

107
Ex. 2-36. 1042/1054, 1, 95-100

1042 violin

 

 

 

 

 

1054 cembalo

‘I

 

In both of these examples from 1042/1054, the original melodic outline is largely lost
in the transcription, again suggesting that Bach’s purpose is not the simple duplication
of musical material from one medium to another, but its transformation into something

individually suited to its new purpose.

III.

The Emancipation of the Cembalo from Thorough-Bass

Of central interest in these concertos is the transformation of the harpsichord from
its traditional role as the backbone of the continuo into a new capacity as the dominant
solo instrument. The previous chapter highlighted some of the more mechanical aspects
of this transforming process, including accomodation to range limitations and adaptations
of string idioms to the keyboard. But for the cembalo to assume its place as principal
solo instrument in an ensemble required a good deal more, including alterations and
additions which were not due to accomodations of violin figuration but which sought to
give the cembalo greater identity and brilliance and thus move it away from its
traditional continuo function.

Bach’s process of transformation as evidenced in these transcriptions is multi-
faceted, and includes attention both to the discant and, more importantly, to the cembalo
bass. While the violin solo material from the models provided most of the musical
material for the cembalo discant, there are frequent alterations or additions which are
obviously intended to make the cembalo part more virtuosic. These include insertion of
passing tones (diminutions), right hand chords when the thematic interest lies elsewhere,
scalar flourishes particularly before cadences, and the addition of new material in places

where the violin solo was silent or played little in the model. This last type is by far the

108

109

most interesting, and is especially well illustrated in the first movement of 1049/1057.
In the first movement of Brandenburg Concerto IV (1049 I, 165 ff.), the solo violin
drops out for twenty bars while the principal musical material and interest is passed to
the two recorders. Since the cembalo discant is derived directly from the solo violin
part, there thus remains no musical material for transfer to the harpsichord’s treble.
Unlike the later concerto of the classic period, in these works the clavier plays
continuously, without any pauses.l The problem faced by Bach, then, was what to do
in the bars in which the original violin solo was tacet. His solution was to invent new
musical material to fill the void, material which forms a counterpoint to the principal
thematic content still played by the recorders. It is also noteworthy that the cembalo
bass duplicates the original cello line and not the more florid basso continuo line of the
model (Brandenburg Concerto IV). By avoiding the rhythmically active continuo bass
of the model, Bach’s new cembalo discant is given room to assume greater importance

in the texture, without competition from the bass (Ex. 3-1).

 

lThis seemingly curious characteristic must not be overlooked. While even the basso
continuo may drop out for extended periods, the cembalo consistently avoids doing so.
The longest "pause,” if it can be so termed, in the cembalo is two-thirds of a bar (in 3/4
meter). That the cembalo is the only instrument of the ensemble to play so continuously
highlights its historical and practical proximity to its other function as basso continuo
instrument. In a sense it might be argued that the cembalo does not become an equal,
fully bonafide solo member of the chamber group until it is freed from the necessity of
playing continuously.

110
Ex. 3-1. 1049/1057, 1, 165-174

1049 cello and continuo

 

1057 cembalo

.__a '——-i--—-‘ -‘—--—- I

.-

I'- .‘1——_‘ AV‘---ri _-1‘-—----=-=
_munu. rl ---———-—————-
—-7 - -

Eur-5

u... ===:I-=-—u .‘.:====-’——

1.1- .r' .——-r--— —-_——_
_- A

 

That this new material in the discant is of musical significance is confirmed by its
reappearance later in the same movement in a parallel passage at mm. 293f. At this
point in the model, the solo violin, while not tacet, plays little of substance and the bass
again is adopted from the original cello part. Comparison of the two parallel passages
(mm. 165f. and 293f.) reveals that the second one begins as does the first, but goes its
own way after four bars. In the autograph. the florid discant is written in tablature
under the cembalo bass beginning at m. 296, the point at which the digression begins.

Thus the reading was arrived at in two stages. Unfortunately, the original version of the

111

discant, written in the treble staff, apparently was erased at some later time and replaced

with the text of the tablature, thus eradicating Bach’s first thoughts on the passage.2

Fashioning the Cembalo Bass

However interesting and informative Bach’s reworking of the cembalo discant
may be in these concertos, it is the fashioning and development of a left hand part which
is the most significant feature in these concertos and perhaps in the establishment of the
genre. As noted earlier, the emancipation of the clavier from its traditional thorough-
bass role was altogether essential if the keyboard concerto was to gain permanence.
Most often, the concertante violin part which served as the source for the cembalo
discant was transferred with very little change, so that the musical material for the right
hand was in a certain sense largely prescribed. For the left hand, however, Bach was
faced with the task of furnishing the transcription with a bass part which had not existed
previously except in the continuo bass of the model. By its very nature, the
thoroughbass character of the original continuo part lacked the musical interest which
distinguished the discant, and if anything this limitation worked against the establishment

of the cembalo as a solo instrument. Nonetheless, Bach still generally chose to use the

 

2Rust stated that Zelter, of the Berlin Singakademie and the owner of the autograph
at one time, took the ”unnecessary pains” to erase the violin figuration and replace it
with the new reading. Based on Rust’s assertion, Siegele repeats this account and
imputes the change also to Zelter. Siegele, Kompositionsweise, 127 fn. 40, and BG 17,
xix. Examination of the relevant measures in P 234 shows that the RH reading is,
indeed, identical with that given in tablature. However, to the extent that can be
determined from microfilm. there appear no signs of erasure such as smudges or redrawn
staff lines, nor are there signs of strike out lines as appears in other passages where
tablature records a second reading. Further and more puzzling, the right hand notation
itself appears to be in the same hand as the surrounding measures.

112

original continuo bass, sometimes quite literally, other times only as a point of departure,
in the formation of a left hand part for the keyboard.

The derivation of this new left hand part from the original bass of the model is
complicated by the simultaneous presence of another bass part, the continuo bass, in the
transcriptions. Both bass parts in the harpsichord versions are descendents from the
original bass of the model, but do not necessarily double each other. There are thus
three bass parts and their relationships to each other to consider: the original bass of the
model; the continuo bass of the transcription; and the cembalo bass of the transcription.
For the four concertos with extant models, these relationships are not uniform but vary
from work to work.

For the Concerto in D Major, 1054, the original bass from the model (1042) is
transferred directly to the continuo bass of 1054 in the opening ritornello.3 The cembalo
bass generally doubles this continuo part, but with occasional interpolated embellishments

such as arpeggiated flourishes during rests in the original bass (Ex. 3-2).

 

3The one minor change is the insertion of passing notes on the second. sixth. and
tenth sixteenths of m. 2. It seems likely that this addition was first made in the cembalo
bass, and then added into the continuo to provide agreement.

113
Ex. 3-2. 1042/1054 1, 1-9 bass lines

(1042 over 1054)

 

In mm. 4-5, the additions to the left hand provide more than just embellishment; without
them, the sparse discant would be left starkly alone during the rests which occur in all
other parts. The left hand arpeggiation thus forms a counterpoint in contrary motion and

strengthens the harpsichord sonority juxtaposed against the tutti (Ex. 3-3).

114

Ex. 3-3. 1054 1, 4—5

 

Subsequent tutti sections in this movement are treated in much the same way, the
cembalo bass and continue bass usually doubling each other with occasional modest
embellishments in the left hand during rests or sustained notes of the continuo (e. g. mm.
8, 29, 32, 49-50, 114-115). In the remaining two movements, this doubling of bass
parts in tuttis is altogether consistent, especially in the third movement where ritornelli
are easily identified by the presence or absence of the continuo bass.

The first solo episode in the opening movement is introduced at the end of the
first ritornello by a brilliant two-octave descending scalar passage which not only extends
the cadence at the end of the ritornello but also signals the entrance of the first solo (Ex.

3-4).

115
Ex. 3-4. 1054, I, 11-13 cembalo

b

 

The parallel cadential extension at m. 34 functions similarly. It is clear from the
autograph that these scalar embellishments were added by Bach after the initial copying
of the manuscript, no doubt as a means of distinguishing the solo character of the

cembalo (Ex. 3-5).

Ex. 3-5. 1054, I, 34-35 P 234

 

 

 

 

 

The formation of a cembalo bass in the solo episodes is in many respects more
telling. Two types of solos, accompanied and unaccompanied, may be distinguished.
In the latter, the original continuo part is transferred to the left hand while the continuo
bass drops out; in accompanied solos, the original continuo is again given to the left
hand, but the continuo either doubles, follows along in simplified form, or remains

silent, as illustrated in Ex. 3-6.

116

Ex. 3-6. 10541, 17—20

--.4- .4
‘—---

—__..
_.

_._.
._._
(I‘mu )nlml

—-. u..-
‘. ’ . 1‘! I

(puam' into)
_—
._- '— _E
.--.—- u- _ __-__‘ .-_ _

___
__.

-r——-----.—- -- ~—
I
p'

_
a

   

I

1054 I. 55-57

_—ii—————_—-_— _—— .—

- LI-V -—r —--.a .-v n_- -——---- a-.. -- ~-——_---- n-a -<

u'__‘ ‘- '-- ~‘——-.a-—- r

1042 I, 17-21 continuo
1042 I. 55—57 continuo

  
   

117

This procedure is especially clear in the third movement of 1054, one of Bach’s most
unusual movements: 3 rondo with five ritornellos all in the same key. This simple
organization makes obvious the alternation of tutti and solo, with the contrast heightened
still further by the continuo’s absence during all solo episodes except the last. The solo
episodes are identifiable, then, as much by the lack of continuo as by the solo character
of the discant. By contrast, in tutti passages the continuo is present and doubles the
cembalo bass.

This same pattern of construction is also evident in the second movement, an
Adagio built over a free ostinato. In the violin concerto model (1042), the ostinato
appears in the continuo; in the transcription, however, it is found in complete form not
in the continuo but in the cembalo bass. The continuo part doubles the cembalo only in
the tuttis, while in solo episodes it drops out altogether or follows the ostinato in outline
only. The extent of the left hand part’s dependence on the original continuo line is
illustrated in mm. 27-30 (Ex. 3-7), where the continuo drops out altogether in the model.
Rather than create a wholly new bass part for the harpsichordist’s left hand during the
continuo’s silence, Bach merely doubles the static repeated eighths of the

viola.

Ex. 3-7. 1054 II, 27—30

IflEfi—Zfi " -
-— ——_. - _. . §
! _—

____________.= __

 

I|
ill
1 l

 

l
l

i'lllllllll
ll

1

:

‘,4——— V

 

 

l
III'

For 1054, then, Bach’s transcription process emerges clearly: the cembalo discant
is derived from the solo violin part with added embellishment and accomodation of
keyboard technique, while the left hand is not newly composed but is the true carrier of
the original continuo part. The continuo part in the transcription basically follows the
left hand part, either by doubling or in a more simplified form.

The Concerto in G Minor (1058) transcribed from the other extant solo violin
concerto (1041) shows both similarities and significant differences in the derivation of
the left hand part. As was the case with 1054. the keyboard discant is derived directly
from the solo violin part of the model, but here with much less alteration of material.
This is due largely to the nature of the violin writing itself: unlike 1042, this concerto
makes no use of repeated notes, broken intervals, arpeggio sequences, or double stops
in the solo part, all of which abound in the E Major concerto.‘ Consequently, the

amount of accomodation to a keyboard idiom is considerably less in 1041/1058. Three

 

4In a certain sense, BWV 1042 shows greater affinity with the style of Vivaldi, while
the formal design and handling of materials in 1041 is of a quite different character.

119
short but parallel solo sections in the first movement of 1058 (mm. 91-94, 106-110, and

129-134) are the only real instances where alterations of the original discant appear.
Unlike the changes in 1042/1054, Bach’s alterations here are not effected because of
violin figurations which are unidiomatic to the keyboard, but rather are undertaken to
make the cembalo discant more sonorous and brilliant by exploiting the upper range of
the harpsichord, or to accomodate a more virtuosic left hand part.

These three solo passages all have similar musical material, and in each the

alteration mainly involves inversion of intervals within the melodic figure (Ex. 3-8).

Ex. 3—8. 1041/1058

1041 I, 91-94 violin

 

1058 I, 91-94 cembalo

 

1041 I, 106-110 violin

 

 

 

 

1058 I, 106-110 cembalo

 

120

1041 I, 129—133 violin

 

 

 

1058 I, 129—133 cembalo

 

While in all three instances the upper reaches of the harpsichord are exploited more
fully, it is more significant that the modifications principally appear to accommodate the
discant to the left hand part. If left unaltered, crossed voices between discant and bass
would result at mm. 91-94, and several consecutive parallel fifths would occur at m.
106. While these problems could have been avoided simply by rewriting the left hand
part, Bach chose rather to give priority to the left hand and accomodate the right hand
to it.

This approach, favoring the cembalo bass, is indicative of the manner in which
this entire work is fashioned, and illustrates a significant difference between this concerto
and 1054. Whereas in 1054 the continuo bass of the model is carried by the left hand
part of the clavier, in 1058 the borrowed bass appears instead in the continuo part of the

transcription, with only minimal changes as necessitated by range.5 This circumstance

 

’Modifications to the continuo part almost without exception involve simple octave
transfers to keep within the available range. In the first movement, examples may be
found at mm. 20, 55, 64-5, 80, 131-35, and 165-66. Additionally, at mm. 31 and 87
octave displacements occur which, while not necessitated by range, avoid the extremities
of the available range. In the autograph, the octave transfer at m. 65 is the result of a
correction, as the original reading followed the registral placement of the model. The
only alteration which does not involve octave transfers is the addition of a single
sixteenth-note non-harmonic tone in m. 55. That this meagre addition is the only change

121

accordingly permits greater freedom in the construction of the cembalo bass than was the
case with BWV 1042/1054, as the left hand part here is freed from dependence on the '
original continua. The cembalo bass of 1058 parallels the continua line most closely in
the opening tutti, where it consistently doubles the continua except for the addition of
a scalar flourish in m. 20. In subsequent ritornellos, however, there is much greater
autonomy, the left hand part often only approximating the melodic outline of the

continua bass, as can be seen in Ex. 3-9.

Ex. 3-9. 1058, I, 51-54

 

In the solo episodes of the model (1041), the continua part is often quite thin, consisting
merely of triadic outlines of the underlying harmony in eighth-note motion. with one or
two bars rest occasionally interpolated. When transcribed for the harpsichord, this thin
bass does not provide sufficient substance to advance the harpsichord as solo instrument.

but it does invite further embellishment. It is no surprise, therefore, to find Bach’s most

 

other than octave displacements corroborates Bach’s strict adherence to the model in this
work.

122

creative reworking and filling out of left hand passagework in these solo episodes. In
the process of developing this passagework, Bach skillfully incorporates motives found '

in the opening ritornello, as illustrated in the first solo episode (Ex. 3-10).

Ex. 3-10. 1058. I, 24-28

   

9 .‘

   
 

l _ ____=._____________

     

 

I-
I
|
|
l
l
l
l
|
I
i
II
I!
1
l

   
  

 

.‘I
I

ll
1
l
l
lll

 

 

  

i

B
g§
{
to

   

III
‘L
J

l
'i
i

 

 
 

1 i
. _ I!
till.
!ll

    

 

 

 

 
 

N
b
p

 

l
'i!
L
l

l
l l
I!
I I
It
I' ‘

 
 

l

l

ll.

    

I _.‘—r—
—-_a - —-_-' _2,
/

 

  
 

 

5 \

 

Although the cembalo bass is embellished significantly in tutti passages throughout the
movement, the most brilliant left hand writing is reserved for solo episodes. The
occasional interpolated rests in the original continua bass line allow the greatest freedom
to the composer in setting the left hand parts, as in the solo episode beginning at m. 84

(Ex. 3-11).

Ex. 3-11. 1058, I, 84-97

 

  

 

, ,—_—_-_ =2 mfir'._—T -— _.—
’) m — m-
y: —=——-‘-=F====—‘-—E=_ ==_—
. . —-— —_=— -_-—— _ —-—_= —
U . ' . _.'
‘ "W I-— 4

.. ___. _ ——_ -.__.__4-_-_'__—~_
in 7— __=_ __ _._~_._ _»_=__-— _= —_=.—
——.,...—, "__— —--...... —-—§

 

 

 

!
j
ill

 

 

 

 

"— —-—‘-

4. .—u .4 ”_.‘. .4 ”mW' _“——"—_
-‘————__-——-—.-u u lr— r-u _——L- -- --.‘v— —
_ _ u—u _.‘ n-— __ Jwe ____-

2
,
\

 

 

I.
ll

 

 

 

 

=-___=
_ ===r====
” ’ 3 w“- "
V g _‘
_ _.____._.._
_— —___.___._______
.3 __ __:=__—r ' —.—_==-=-_.___
- ——-' —- _.7 ‘—
—. _— -=5
.' _:=-_— __ _.__——_-.——_=_—-—-—
_ ———= ==_=-_—==.=
a”: ‘ - ' .-;—__—-———_'_.s.=rz_-=:_ ' = - -r4:a:e.=e.=::é_.e-é5:_- -u
6“ -... .. "-E"-'— ___._r. m -_ n- ......
a... "I!— .- -
3.. _ -5;— , -‘-_-_- _- ' =97 _. -39"; __
l ' _‘ _7_- _I ‘ — _.’

 

 

for the second and third movements, but not to an extent which alters the principles
found in the opening movement. For the Andante, Bach simply doubles the lowest
sounding voice, which is the continua part in the tuttis and the viola part in solos where
the continua drops out. This latter process of creating a bass part by doubling the viola
is exactly the same procedure as found in mm. 27—31 of the second movement of 1054

cited earlier. In the final movement, Bach again relies upon the original continua bass

124

more closely than in the opening movement for the derivation of a left hand part, but it
is clear nonetheless that this borrowed bass is placed not in the cembalo bass but in the
continua. The left hand part then, while following the outline of the continua more
closely than in the first movement, still finds many occasions for added embellishment
while the continua remains true to the original bass. All manner of interesting ornament
is brought to bear in the left hand, including arpeggiation and sequential repetition of

newly devised figures which go far beyond the sketchy continua bass (Ex. 3-12).

Ex. 3-12. 1058, III, 73-80 continua and cembalo

 

Accomodation of violinistic figuration in the cembalo discant summons brilliant keyboard
effects in the left hand at mm. 82 and following, succeeded by a three-bar culminating

passage over a prolonged dominant pedal point in the continua (Ex. 3-13).

125

Ex. 3—13. 1058, III, 83-90 continua and cembalo

 

It becomes obvious from the foregoing examples that throughout this concerto the
real carrier of the original bass is the continua part in the transcription, and that the left
hand part, while not totally divorced from the continua line. is sufficiently independent
of it to be allowed greater freedom and individuality. In other words, in this concerto
the solo harpsichord departs more fully from its traditional role of thorough-bass
instrument. It also now should be apparent why Bach copied in the upper string parts
in the autograph first, leaving not only the cembalo but also the continua to be worked
out later. The varying and often interdependent relationships between left hand and
continua makes it clear that these two parts had to be worked out jointly at the same

time.

Special Considerations in BWV 1057

The Concerto in F Major (1057), modeled after the fourth Brandenburg concerto,

differs on several levels from the other harpsichord concertos contained in the autograph.

126

The sheer length of the opening movement--427 bars, including the 82 bar opening
ritornello--places the work in an altogether different category from 1054, 1058, and
1062, all of which were modeled after the considerably shorter violin concertos.“
Although the violin principale is the main solo instrument in 1049 (transferred to the
discant of 1057), the presence of two recorders significantly alters not only the timbre
but also the complex relationships between parts, since thematic material is sometimes
shared between the recorders and solo violin.

As Adolf Aber noted many years ago, this concerto has the character of a
concerto grosso for three instrumental groups: clavier, two recorders, and tutti, but with
the clavier made more brilliant than the violin part of the model.7 The importance of
the recorders to this work is evident already in the long opening tutti, where principal
and secondary themes are more often played by the recorders than by the ripieno
strings.8 For much of the opening ritornello, this circumstance results in the cembalo
merely filling in with chords rather than doubling the solo violin as occurs in BWV 1054
and 1058.9 A somewhat remarkable feature of 1057 I is the occasional transfer of
musical material from the solo violin of the model to the left rather than the right hand

of the cembalo.10 In these concertos the appearance of solo thematic material in the

 

“The first movement of 1041 has 171 bars, 1042 I contains 174 bars, and 1043 I is
the shortest at only 88 bars.

7Adolf Aber, ”Studien zu J. S. Bachs Klavierkonzerten," B] x (1913), 8.
“See mm. 1-13, 25-29, 35-47, 59-62, 65-69, and 75-82.
9Examples may be found at mm. 1-12, 25-28, 31-35, and 59-68.

10See mm. 63-69, 125-139, 263-270, 311-328, and 407-411. The first and last of
these are identical, while the other three passages are related.

127

cembalo bass is a rare occurrence in non-fugal movements. Nowhere in 1054 or 1058
does material derived from the solo violin in the model appear in the bass (the third
movement of 1058 is fugal, and accordingly is outside this cohsideration).“ Hence the
several appearances in 1057 of solo material in the cembalo bass are all the more
conspicuous and provide a more soloistic function for the left hand, moving it further
from the traditional continua role. In a number of these passages, such as mm. 125—139,
the principal thematic material is played by the recorders while the left hand of the
cembalo carries the countermelody which had been played by the solo violin in 1049,

but now sounding an octave lower (Ex. 3-14).12

 

1‘In 1058 I brief motives drawn from the principal thematic material of the opening
tutti do appear sporadically in the bass, as at mm. 26-27, 67-69, and 86-88. These,
however, are the product of embellishment and do not represent a transfer of the solo
violin part to the left hand.

12This countermelody is ingeniously derived from the opening tutti (mm. 14, 16, 18-
22).

128

Ex. 3-14. 1049/1057, 1. 125-130

1049 violin and recorders

 

1057

 
        

.—
-—v '1 n-.- —1 .——.- -‘ .a n-‘—.- '1: _-
‘ — _—
.‘ 2“ —‘--—---—-_‘-—==
-u_———_————-—__ ,2 _—

       
   
  

  
 
  

I
I
I
I
i
I
1
I

 

 

 

 

 
 
      
       

g H 7 —_-—

_— - _.un r-u—r—
————- ——*r- r. ll- 1 r. L-J

_— —' —_'
_ _ r— !
=——=———- —-—_
_— u——_—
- _— -——

I
r.

. ,M
, “IN
MI
Illfl

            
  

I
I
1|_
1
I
I
I
I

ll

_v-

I
l
I
i

-. 7
’7

I
I
I
I
II

“i
i
i:
I
l
i
l'
l

    
    

i
"i!

ii
i
i

 
       
 

l
i

~_'-

J

|
Ill

   

-———-—

  

The fundamental question again is the relationship between the new cembalo bass and
the continua parts in both the transcription and the model. In this particular concerto
these relationships are more complex than in 1054 or 1058 due to the expanded scoring
and the fact that in 1049 Bach actually wrote three bass parts, for Violoncello, violone.

and continua. As might be expected, the violone mostly doubles the cello and/or

129

continua bass, but not always; there are passages, particularly in the first movement,
where the violone does not simply double the other basses, a circumstance that has
import for the distribution of parts in the transcription. Although the three bass parts
most commonly double each other, other combinations also appear, so that a certain
degree of independence between parts results. One of the most interesting passages in
this regard comes at mm. 298-310 of the first movement, where the three basses of the
Brandenburg model are not identical. In the harpsichord concerto arrangement, the new
continua line is taken from the violone of the model while the cembalo bass is the former
cello line; the original continua, which is a slightly embellished version of the cello part,

is not used (Ex. 3-15).

Ex. 3-15. 1049/1057 1, 298-305

1049 basses

 

1057 continua and cembalo

 

130

The manner in which Bach accommodates three bass parts from the model to two
bass parts in the transcription is noteworthy. In actuality, all three bass parts from 1049
are retained in 1057 within the two written bass parts, since the continua part is played
at times by either or both the violone and Violoncello. The marking at the outset of the
BG edition violone e Violoncello is lacking in the autograph, but the indications violone
and Violoncello do appear periodically in the continua part of the manuscript, suggesting
that Bach intended both instruments to be involved in the continua part.13 This double
instrumentation appears to be highly exceptional, for in none of the other concertos,
whether for solo or multiple keyboards, does this practice of differentiating between cello
and violone within the continua part occur. In all the multi-keyboard concertos, the
NBA assumes the use of both cello and violone for the continua part, even though
textual evidence to substantiate such an assumption is lacking. At the top of the
autograph of the Concerto for 2 Claviers in C Minor (1062) appears the title in Bach’s
own hand Concerto a due Clavicembali obligati. 2 Violini, Viola e Violoncello di Bach.
Despite this ascription, the NBA still lists the continua instrumentation as Violoncello and
violone, albeit with the latter in italics. For 1060 the autograph is not extant, but there
are several fairly reliable copies, including one in Altnickol’s hand which lists the ripieno
instrumentarium as Violino 1 / Violino 2 / Viola / con / Violon / d. S.J.S. Bach.l4

Perhaps it was this reference to Violon that led the NBA editors to assume the vialone’s

 

13These indications appear in P234 as follows: in I, violone is marked at I, 103,
125, 311, and III, 129; Violoncello at I, 98. 120, 258. 311 and III, 101. The designation
Violone e Violoncello encountered in the BG (I 1, 103, 137, 249, 271, 323; III 14, 12)
appears nowhere in P 234.

l“Mus. ms. Bach St 136. See NBA KB VII/5, 13.

131

part in the instrumentation of all these concertos.” However, other copies considered
to be as close to the original autographs as Altnickol’s cite only Basso continua or
Continua for the part.” Hence, Altnickol’s reference to Violon ought not to be taken
as definitive. In the surviving copies of the concertos for three harpsichords (autographs
not extant), the bass is labeled by the generic Continua, Basso continua, Bass, or the
like, the one exception being a copy from the nineteenth-century which indicates
Violoncello e Violone.l7 Similarly, manuscript copies of 1065 for four claviers make
no reference to violone, and for the solo concertos (1052-1059), Bach’s own titles from
the autograph refer simply to continua. Thus it may be stated that 1057 is unique among
these concertos in its reference to the violone in the continua. It also follows that this
exceptional use of violone in 1057 in all likelihood owes its provenance to the larger
instrumentation of the Brandenburg IV model, thus corroborating Bach’s close
dependence on his model.

Many questions nonetheless remain about the inclusion of the violone in this
concerto. Laurence Dreyfus has suggested that the regular participation of the violone
in performances at Leipzig is particularly difficult to establish, since the violone was

likely to be one of the first instruments omitted when performing forces were limited.18

 

1“This assumption that the continua includes violone is to be found also in the NBA
edition of the concertos for three and four harpsichords. See NBA VII/6.

1“NBA KB VII/5, 14-15, 34.

17Parts copy H 803. NBA KB VII/6, 15. The late date of this source. coupled with
the existence of other earlier sources which do not mention the violone, casts doubt on
the reliability of this ascription.

18Laurence Dreyfus. Bach ’s Continua Group, 156.

132

The list of necessary players enumerated in Bach’s famous ”Memorandum for a Well—
Appointed Church Music” of 1730 includes a vacant place for violone,19 though as
Bach also noted these vacancies could sometimes be filled by students from the
University, some of whom no doubt also participated in the collegium and thus were
familiar to Bach. The list of extant works from the collegium years which call for
violone is relatively small, but nonetheless opens the possibility of the instrument’s
availability when necessary: Cantatas 182, 245, 226, 42, 62, 214, 91, 100, and 208a.
The last of these works, Cantata 208a, also includes in its instrumentarium two
recorders, Violoncello, violone, and continua, which is notably similar to the
requirements of both 1049 and 1057.20

From the sources, Dreyfus concludes that three different violone types were used
by Bach at different times in his compositional career: a small six-string violone in G
that sounded at pitch; a larger six-string contrabass violone in D sounding an octave
lower; and a four-string Violone grosso, also sounding at 16’, which reached to CC. For
the Leipzig works, Dreyfus holds that the violone in use was a 16’ instrument. thus
sounding an octave lower than written.” In a larger scored work, this low fundament
is more conceivable, but its suitability in 1057 might be questioned. Since the violone
and Violoncello double each other in 1057 with few exceptions, the result would be a

reinforcing bass under the harpsichord and recorders. In a number of instances,

 

19The Bach Reader, 122.

”Cantata 208a also includes oboes, oboe da caccia, horns, bassoons, and the usual
strings.

21Dreyfus, Bach ’s Continua Group, 142, 158, 161.

133

however, this results in a two-octave distance between cembalo bass and continua
(violone) if the latter sounds an octave lower than written. This circumstance may be I
seen at 1057, 1 mm. 125-133, where the cello is in the same octave as the cembalo bass
while the violone is notated divisi an octave below. If a 16’ instrument, two octaves

will separate the violone from all parts above it (Ex. 3-16).

Ex. 3-16. 1057 I, 124-130 continua and cembalo

 

Dreyfus has noted in 1049 that the violone part consistently avoids C, this avoidance
being particularly conspicuous in passages where the violone parallels the cello, and
accordingly concludes that the violone called for in Brandenburg IV was a 16—foot
instrument with a low D string.” In the transcription (1057), however, this same
instrument cannot be intended, for the downward transposition of key results in
numerous occurrences of C in the violone part, suggesting that the violone in 1057 is the
violone grosso with a low C-string as used in Brandenburg Concerto I, or is the smaller
violone in G used in Brandenburg Concertos II and VI and which sounds at pitch, or that
neither instrument is necessarily implied but that the violone designations in 1057 mean

something else.

 

22Ibid., 144.

134

Of these three possibilities, the first, the violone grosso with low C. seems least
likely owing to the two-octave distance between bass parts that inevitably results. A
further objection to this instrument is its upper range. Tuned an octave below the cello,
CC, GG. D, A, its upper limit would have been c’. The cello/violone part in 1057
reaches (1’ at several points, and e‘” once in the second movement. By contrast, the
smaller violone in G, sounding at 8’ pitch, could conceivably be the intended instrument;
Walther indicated that the upward ambitus of this instrument extended to d’ or e’.
However, while Dreyfus has been able to support the hypothesis that the violone in G
likely was used in works from Weimar and Cothen, its application in Leipzig is
unsubstantiated; on the contrary, the musical sources tend to suggest that the Leipzig
violone played at the lower octave.”

This leaves the third possibility: that perhaps the violone designations indicate
something else. Table 4-1 lists the occurrences of the labels violone and Violoncello in

the BG edition and the autograph.

 

23Ibid., 160-161.

135

 

Measure Bach Gesellschaft Autograph
I l Violone e Violoncello
98 Violoncello Violoncello
103 Violone e Violoncello Violone
120 Violoncello Violoncello
125 Violone Violone
137 Violone e Violoncello
244 Violoncello
249 Violone e Violoncello
258 Violoncello Violoncello
271 Violone e Violoncello
311 Violoncello / Violone Violoncello/(Violone)”
323 Violone e Violoncello
III 14 Violone e Violoncello
101 Violoncello Violoncello”
129 Violone e Violoncello Violone

Table 4-1. Violone and Violoncello indications in 1057

It is clear from the above that the indication Violone e Violoncello was added into the
score by the editors of the BG; nowhere in the autograph does this reference to both
instruments together occur. Further, and more important, comparison of these
instrument designations in 1057 with the parallel passages in 1049 suggests a possible
purpose or meaning in their use: in every instance where Violoncello appears in P 234,
the violone has dropped out at the corresponding passage in 1049, leaving the cello and
continua to play the bass. Similarly, when the violone reenters a number of measures
later in the model, the indication Violone appears in P 234, thus marking the return of

the violone doubling the cello in 1049. The result is paired markings of the exit and

 

2‘ The two instruments are divisi here. In P 234 only ”Violoncello" appears for the
upper part, but the passage is parallel to that found at mm. 125 f. , where the lower part
is marked "Violone” while the upper part continues the cello part. This same division
of the two instruments clearly is intended at mm. 311 f.

”In the autograph, this marking is very light, almost illegible.

136
reentry of the violone in the Brandenburg model: I 98-103, 120-125, and 111 101-129.

The only seeming exception to this arrangement is at I 258, where the violone drops out
in 1049 until measure 271. At measure 271 the full tutti reenters, however, and though
not expressly marked Violone in P 234, the return of the violone may be assumed.”
In this light, it is noteworthy that the paired markings in the BG at I 244-249, absent
from the autograph, are unnecessary and incorrect, since at this place in 1049 the violone
does not drop out and hence a reduction to Violoncello is not warranted. Thus the
instrument designations in the continua part of P 234 may be more graphological
signposts to the origins of the part rather than definitive prescriptions of the
instrumentation, particularly since two bass parts are encompassed in a single staff.
Admittedly such a hypothesis does not answer the problem of the two divisi
passages (I 125-133 and 311-319) which comprise but 6 bars of a total of 742 in the
concerto. This very limited occurrence prompts a question: to what extent should these
6 bars impose an interpretation of the instrumentation, particularly when one considers
that the violone is not so much as mentioned anywhere in the title of this or any other
concerto in P 234. For 1057, the title at the head of the score lists the instrumentation
as ”Concerto e Cembalo certato, due Flauti a bee. due Violini, Viola e Cont." By
contrast, the title of 1049 is far more descriptive: ”Concerto 4‘o a Violino Prencipale.
due Fiauti d’Echo. due Violini, una Viola e Violone in Ripieno, Violoncello e

Continua.” It is notable that in two other score copies of Brandenburg IV this exact

 

2“In P 234, there is the appearance of a light shadow or possible erasure at m. 271.
It is difficult to determine more from the microfilm.

137

same title and listing of instrumentation are retained,” suggesting that when copies of
a work were made it was customary to preserve the title exactly as it appeared in the
original. The lack of this same instrumentarium in the title of 1057 thus becomes all the
more conspicuous.

In each instance where the violone and cello are divisi at an octave’s distance in
1057. the same octave spacing appears in 1049, the cello taking the upper pitch while
the violone doubles the continua part at the lower octave. This whole question might
not be an issue at all if the cembalo bass simply replicated the continua part as found in
1049; but in both divisi passages, the cembalo bass is given the original solo violin part
(transposed down an octave). The left hand being thus occupied, the original continua

bass part is essentially absent except for its preservation in the violone part (Ex. 3-17).

Ex. 3-17. 1049/10571, 125-133

1049 basses

 

 

2“’Am. B. 77 and Mus. ms. Bach P 259. See NBA KB VII/2 87-88.

138

The seemingly obligatory inclusion of these bass pitches from the violone and
continua parts of 1049 illustrates the very dependent relationship between 1057 and its '
model, and also raises questions whether the presence of these divisi pitches are not
merely the product of compositional process or transcription, without regard to their
practicality or performability. A literal interpretation of every detail of Bach’s score
may be questioned at other points as well, particularly in the light of performance
practicalities. At m. 325 of the same movement, the left hand of the cembalo sustains
two a’s an octave apart for over four bars. The upper a serves as the extension of the
theme begun two bars earlier (played by the violino principale in 1049), while the lower
A is derived directly from the unison bass parts of 1049. The right hand is preoccupied

with brilliant new material not found in the model (Ex. 3—18).

Ex. 3-18. 1057 I, 323-329

 

Inasmuch as such long held notes dissipate quickly on the harpsichord, Bach adds the
customary trill signs over each a, a practice which is consistent in all these concertos.
In the present instance, however, the result is an interesting technical improbability, for
it is physically impossible to play with the same hand two simultaneous trills an octave
apart on the keyboard. Indeed, even playing a single trill on the upper a (with the upper

neighbor b) while only holding the lower A may not be technically possible for all

139

players. Thus the trill signs cannot be taken literally, but may be considered a
graphological reference to a compositional convention rather than a mandate for
performance realization.

Generally, the basso continua in 1057 is taken directly from the violone rather
than the continua part of the model.” Conversely, the original continua bass of the
Brandenburg model generally is transferred to the left hand of the cembalo. In this
respect, this concerto shows some affinity with 1054. However, there are also
significant differences which set 1057 apart. Whereas in 1054 the continua bass of the
model is virtually always present in the left hand either in original or embellished form
and doubled by the new continua part, in 1057 the original basso continua occasionally
migrates to the continua part of the transcription, particularly when the cembalo left hand
is needed to assist in solo figuration. Thus in the most brilliant solo episodes, such as
at mm. 98-103, 119-125, and 191-209, both hands are called upon to negotiate rapid
scalar passages, requiring the left hand to relinquish its bass-playing role. In such

instances, the original continua bass is then transferred to the continua (Ex. 3-19).

 

2“Comparison of the violone part of 1049 with the continua part of 1057 verifies this
to be the case. In instances where the violone drops out in the model, the continua is
also absent in 1057 (e.g. 165 ff).

140
Ex. 3-19. 1057 I, 193-203 continua and cembalo

 

In a somewhat similar manner, the left hand occasionally takes over the solo violin part
of the model, transposed down one or two octaves, requiring that the continua part again

carry the borrowed bass (Ex. 3-20).

Ex. 3-20. 1057 I, 125-132 continua and cembalo

l

  

 

 

I“ _———_
I--—__—_—

Compared with 1054, the distinction of this concerto rests in the significantly greater
importance of the left hand part. as evidenced by the appearance of solo thematic

material,” participation in brilliant solo figuration divided between the hands,30 and

 

2"Examples include] 63-4, 125-137, 240-243, 249-250, 263-271, and 311-325. The
fugal manner of the third movement results in thematic appearances at mm. 15-19, 67-
71, 135-139,, 167-171, 207—211, and 237-241.

141

in the more rhythmically active figures in tuttis.31 Despite the left hand’s general
dependence on the continua part of the model, the solo instrument is elevated to a

greater degree and thus advances the whole work in the direction of a true solo concerto.

Transcription to Two Keyboards: BWV 1062

The fourth and final work from the group of concertos with extant models is the
Concerto for two harpsichords in C Minor, BWV 1062, arranged from the more famous
Concerto for two violins in D Minor, BWV 1043. As already noted, this work appears
to date from 1736, near the end of Bach’s first period of association with the Collegium.
Most fortunately, the autograph score of this concerto survives in Mus. ms. Bach P 612,
the only known Bach double manuscript containing two independent works of different
origin and instrumentation.” As the manuscript has a cover with the titles of both
works in the hand of Philip Emanuel, probably written around 1767 during Emanuel’s
Hamburg years, it is likely that the score was at some time in his ownership.”

Although this concerto is arranged for two harpsichords rather than one, the process of

 

”Examples may be found at I 83-102, 105-125, 191-209, 244-248, and 258-262; III
101-120.

31I 20-23, 41-47, 53—56, 271-275, and similar passages throughout the movement.

32While the Concerto in C Minor, BWV 1062, is the principal work in the
manuscript, the Flute Sonata in A Major, BWV 1032, occupies the remaining bottom
staves on each page of the manuscript.

33NBA KB VII/5, 60. A facsimile edition of Mus. ms. Bach P 612, edited by Hans
Joachim Schulze, has been published in Documenta Musicologica. Zweite Reihe:
Handschriften-Faksimiles X, Johann Sebastian Bach: Konzert C-dur fiir zwei Cembali
und Streichorchester, BWV 1062 / Sonata A-dur fiir F late und Cembalo BWV 1032
(Kassel: Barenreiter, 1980).

142

transferral from model to transcription is very similar to that found in 1054 and 1058.
The presence of many corrections in the autograph, particularly in the left hand parts of
both cembali, indicate that this autograph also was a working score rather than a
presentation copy. Whether such changes came as a result of Collegium rehearsals and
performances, as speculated for 1054 and 1058, or were worked out at the time of
transcription is impossible to ascertain. Schulze holds that the clavier version was
derived directly from the violin version without any intermediary.” As in the solo
concertos, the original solo violin parts are transferred to the harpsichord discants, while
the left hand parts are adapted freely from the continua bass of the model. The ripieno
strings follow the model virtually note for note.”

As in the concertos previously considered, it appears from the autograph that
Bach first wrote out the ripieno string parts, followed by the continua and keyboard
parts, with the left hand parts of the cembali probably worked out last with still further
alterations. Parallel to the autograph of 1054, this sequence is betrayed by frequent
instances of sixteenth-note passages spilling over barlines, both in the left hand parts as
well as in the continua bass, often with the barline distorted so as to accomodate the

additional notation. Even when sixteenth-notes are present in the upper strings, they fit

 

3"Ibid., 15.

3“The one notable exception appears in the first movement, mm. 22-23 and 26-27,
where Bach apparently first copied out the same parts for ripieno strings as in 1043, but
subsequently "corrected” them to a different reading which does little more than change
the distribution of parts: in m. 23 the lst violin part of 1043 becomes the viola part of
1062, the 2nd violin part is moved to lst violin, and the viola of 1043 is taken over by
the 2nd violin in the transcription. The alterations are clearly visible in the autograph,
but no such changes appear in parallel passages later in the movement. There is no
obvious or apparent reason for these changes.

143

comfortably within the barlines with rare exceptions. In more limited instances,
sixteenth-note passages in right-hand parts may also breach the barline occasionally,
indicating that not only the bass but also the keyboard discant was copied out after the
ripieno string parts.

While the transfer of two solo violin parts to two harpsichord discants was
relatively simple, the creation of not one but two cembalo bass parts from only one
continua part is necessarily more complex. Generally, the continua part of the model
is transferred to the continua part in the transcription. with the left hand parts derived
to a greater or lesser degree from this same continua line, particularly in tutti sections.

In order to avoid overuse of doubling, octave placement is sometimes altered (Ex. 3-21).

Ex. 3-21. 1062 1, 5-6 cembali

 

In solo episodes, the continua bass of the model is transferred to the continua part of the
transcription, but the left hand parts are much freer in their relationship to the continua
bass than in the tutti sections, often providing a much more brilliant accompaniment to
the clavier discants. In Ex. 3-22, the first solo entrances of the harpsichords at mm. 22

and following illustrate.

Ex. 3—22. 1062 I, 22—29 cembali

I‘v-_- _ H-u— — -——-—————--—.-——— --———-————————-.——
um~——- -—-—"---—-———-- ——-————_—-——-—— —
—-‘——————-—.-—————-_._——————-—. ——- -————

- —— _—

 

In effect, this differentiation in the handling of left hand parts in tutti versus solo
passages is so distinct as to virtually signal the return of tutti ritornellos: when the
clavier basses move freely, the passage is usually an accompanied solo, but when the
keyboard basses simply double the continua, the passage is almost always a ritornello.
Thus the harpsichords fulfill a double function: in tuttis, the keyboard discants double
the ripieno strings while the left hand parts serve as continua bass; but in solo episodes,

both parts, right and left hand, assume more brilliant solo roles. This pattern is most

145

evident in the first movement, but is also recognizable in the finale. The second
movement, lacking the sharp contrast of tutti and solo sections, finds the keyboard bass
largely doubling the continua line. Thus 1062 is similar to the workings of 1058, for
in both works the continua bass from the model is transferred to the continua of the
transcription, while the clavier basses are freed to function more independently. By
contrast, in 1054 and, to a slightly lesser extent in 1057, the left hand parts largely carry

the original continua throughout.

Added Harmonic Filler

Before attempting to draw any conclusions concerning these left hand parts in the
concertos transcribed from models, some observations concerning the addition of chordal
"filler” need to be considered. Traditionally, the harpsichord’s role as continua
instrument was fulfilled by the keyboard player supplying a chordal realization of the
harmony, and for this function the harpsichord was intrinsically well suited. It might be
considered, then, to what extent the harpsichord parts in these concertos under discussion
either continue this function or are emancipated from it. At least for the four concertos
considered here, there is not one clear answer. At the one extreme is the F major
concerto, 1057, in which extensive chordal filling-in by the harpsichord is common.
While this chordal filler appears most frequently in tutti passages, it is not altogether
absent in solo episodes. Further, this harmonic filler is not limited to only the right hand
part. but may be added into either discant or bass. While some of these passages are no
more than doublings of ripieno parts, occasionally the harmonic filler is not derived by

doubling. The opening of the first tutti well illustrates the latter possibility (Ex. 3-23).

146

Ex. 3-23. 10571, 1-10

Fltmlo I.

«Hire a but.)

Flautlo I I.

"like a tad

Violino I.
Violino ".

Viola.

\‘Iulau c Viola-"Ila.

_-

Cuttlitmo.

(’r-tn hula.

 

Similarly in the second movement extensive chordal filler appears in the cembalo discant

during tuttis such that a very discernible alternation between solo and tutti becomes

evident (Ex. 3-24).

147

Ex. 3-24. 1057 11. 9-15

_ , f‘

—— __ n
_—
. . _—
_ _—
—— “_—
— — _—

-—-—_—— m — n _—
—- —'-—--— — —— ——m¢ "a
_- — _.- --- —.—— _—|.__-

 

Even the finale, despite its fugal design, contains episodes in which the cembalo fills in
chordal harmony.” As noted earlier, this concerto is unique among these arrangements
for being closer to concerto grosso than solo concerto, and this fact influences the"
function of the harpsichord to a considerable degree. Given the amount of chordal filler,
the harpsichord here more closely reflects the traditional continua function than is true
in the other concertos.

By contrast, 1062 for two harpsichords may be regarded as the opposite
exemplar, with no reliance on filled in chordal harmony whatsoever. In certain passages
where the original bass is thin, Bach occasionally added a broken octave or arpeggio

rather than black chords, as in the following example (Ex. 3-25).

 

”See especially mm. 67-86 and 138-152.

148

Ex. 3-25. 1062 III, 12-14

 

The total absence of any chordal filler in this concerto is remarkable in its consistency:
even the final tutti cadence avoids any use of filled-out harmony. Hence, at least in this
regard, the harpsichords in this concerto are more distant from the traditional continua
function and mare closely approach a true solo role.

The two remaining concertos, 1054 and 1058, demonstrate a middle ground.
While some chordal filler may be found in both works, there is considerably less use of
it here than in 1057, and usually it is of a different kind. In 1054, added chords are
found only in the first movement, and then only in tuttis, and typically consist of little

more than doublings of ripieno string parts in parallel thirds (Ex. 3-26).37

 

3"Similar passages occur at mm. 1-2, 20-21, 38-39, 70-71, 73-74, and in parallel
passages in the da capo.

149
Ex. 3—26. 1054 I, 15-16

__ —
.. — — ——————-———.—-—-

. ' ' — -_.._-—=—-— -—=——-——_—.__:
.. . .—_.-— —-_r-— _- _ — r-_-

:‘"“"- _u_‘—"-"

 

The sole exception is in the first half of m. 35. where three added chords reinforce the
harmony but are not the product of part doubling. In 1058, added chords in the cembalo
appear in tutti sections, especially at cadences.” Such use is rather unexceptianal, and
is usually the consequence of doubling. By contrast, considerable use of chordal filler
is found in the second movement of this same work, where it marks the clear alternation
of tutti and solo episodes so characteristic of this movement. In this regard. this

movement bears similarity to the second movement of 1057.

Conclusions

Some conclusions. then, may now be drawn from these four concertos with extant
models. Two approaches to the setting of left hand parts may be distinguished: on the
one hand are two works (1054, 1057) in which the left hand carries the original continua

from the model, the new continua line either doubling or simplifying it; on the other

 

3"Instances of chordal filler at cadences occur in mm. 17, 23, 72, 83, 124-125. 159,
and 170. Most of these are parallel passages.

150

hand are examples (1058, 1062) in which the borrowed continua part is transferred
directly to the continua part of the new keyboard arrangement, thus freeing the left hand
parts for a more independent, soloistic role. Further, if the solo role of the clavier is
acknowledged to be distinctly separate from the continua function, then the considerable
number of continua-like chords added into 1057 suggests that the harpsichord combines
the functions of continua and solo instrument in this work. Additionally, the clavier bass
is basically the same as the continua bass of the model, since this concerto represents the
type of transcription in which the original continua part is transferred virtually verbatim
to the new left hand part. The combination of a model-dependent bass and considerable
chordal filler in the same work suggests that this work is less progressive, at least when
compared with concertos in which the left hand part is free to assume a more soloistic
role. It is notable that of these four concertos with extant models, 1057 is the only one
for which the model can be proven to stem from Cathen. That the transcription thus was
fashioned at least fifteen years after the date of composition of the model may help
explain why 1057 is unique in including new musical material not found in the original
Brandenburg version of the work. Perhaps after so long a time interval, Bach rethought
the musical ideas of the original work, and accordingly was more disposed to incorporate
changes. At the same time, however, these changes did not go so far as to free the
cembalo left hand part from the original bass and thereby capture a more soloistic role
for the harpsichord. On the other hand, the more conservative approach to the
harpsichord in this concerto may be due to the rather unusual nature of this work, which
approaches concerto grosso. If the organization and layout of the autograph is taken to

indicate that at least the first six concertos were written about the same time, then there

151

would seem to be little other explanation. One might speculate that 1057 could have
been transcribed earlier and was copied into the autograph sometime in the late 1730’s;
but the presence of so many corrections in the score suggests otherwise: that P 234
represents a first working copy rather than a textual copy from a previous arrangement.

While 1054 and 1058 both use some chordal filler, it is of a different kind than
that found in 1057. In 1054 and 1058, the added chords usually are derived through a
simple doubling of string parts. The violin concertos 1041 and 1042 which served as
models for both of these arrangements have been credited to the C'o'then period, even
though no documentation to support such assignment exists.” Wolff has suggested that
with the absence of extant composing score sources, the existence of autograph materials
datable to Leipzig, and a style-critical appraisal which finds BWV 1041 and especially
1043 consonant with Bach’s style in Leipzig, there may be grounds to support a Leipzig
dating for these works."0 While no clear conclusions can be definitively drawn, it
nonetheless remains that in both 1058 and 1062 (the arrangements of 1041 and 1043
respectively), the cembalo left hand parts are more independent of the borrowed bass.
When this observation is coupled with the total absence of chordal filler in BWV 1062,
the inevitable speculation emerges that these works somehow represent a later,
culminating stage of development in Bach’s approach to the concerto.

It becomes evident that there is not one single transcription pattern that emerges

 

39Extant source materials for 1041 and 1042 can be dated only to c. 1730 and 1760
respectively. NBA KB VII/3, 11, 21-22. Most scholars have defended earlier, pre-
Leipzig dating solely on the basis of style criticism.

4° Wolff, ”Chamber Music," 172-174. As Wolff notes well, there are significant
differences between 1042 and 1041, the latter displaying much more developed and
mature characteristics.

152

from these four concertos; at best, there is a sense of a transitional or developmental
process in varying stages of refinement. Despite some continuing use of added harmonic
filler, the cembalo parts in these works, when compared with their source models,
evidence to varying degrees the development of independence and figuration idiomatic
to a keyboard instrument, thus moving the cembalo away from its continua heritage.
Once only subservient to the ripieno, it now asserts dominance on several levels:
thematic, timbrel, and virtuosic. Though still strongly tied to its contrapuntal heritage,

the cembalo here presumes to mark out new territory.

IV.

Concertos Without Models:
Comparisons with Parallel Versions

While the remaining concertos in P234 have no extant models, parallel settings
of several of them do survive in certain church cantatas with elaborate obbligato organ
parts, the latter paralleling the cembalo parts in the concertos. Most of these cantata
movements are extended instrumental sinfonias scored for strings and organ and which
clearly bear the imprint of concerto design. The relevant concertos, with their parallel
versions, are as follows:

Concerto in D Minor, BWV 1052

I. Allegro Cantata 146, Sinfonia
II. Adagio Cantata 146, II
III. Allegro Cantata 188, Sinfonia1

Concerto in E Major, BWV 1053

I. [Allegro] Cantata 169, Sinfonia
II. Siciliano Cantata 169, V. Aria
III. Allegro Cantata 49, Sinfonia

 

1The parallel version of this movement, a Sinfonia to Cantata 188, unfortunately does
not survive; that it once existed in this form is attested to by references in two
fragmentary sources for the cantata, one now located at the Bibliotheque du
Conservatoire de Musique in Paris, the other at the Bibliothek der Gesellschaft der
Musikfreunde in Vienna. Both fragments consist only of a folio page, but refer to the
use of an organ concerto as introduction to the cantata, with the incipit of the D Minor
Concerto shown to identify the intended work. See Siegele, Kompositionsweise, 113-
114, and BG 37, xxxvii-xxxix.

153

154

Concerto in F Minor, BWV 1056
I. [Allegro]
II. Largo Cantata 156, Sinfonia
III. Presto
Additionally, the 9—bar fragment of the Concerto in D Minor, BWV 1059, which appears

at the end of P 234 is paralleled by the first movement of Cantata 35.

Comparisons of Variant Readings

The existence of parallel yet quite different versions inevitably prompts questions
about the relationships between such variant settings, including the interdependence of
each version. While the particular conditions of each concerto vary, it appears that in
each instance the concerto was not derived from the cantata, nor the cantata from the
concerto, but both hark back to an earlier, now lost model. The issues which lead to
such a conclusion can be illustrated rather clearly using the Concerto in E Major (1053)
as an example. When preparing the BG edition, Wilhelm Rust assumed that this
concerto was transcribed directly from the Sinfonias to Cantatas 169 and 49.2 The fact
that both cantatas stem from 1726 while the 1053 autograph dates to the late 1730’s
would seemingly support such a presumption. Spitta, however, while allowing that the
concerto could not have preceded the cantata, also concluded that the alto’s melodic line
in the Aria (#5) of Cantata 169 (parallel to the second movement of the concerto) was
not a part of the movement originally but represented a later addition not found in the

concerto.3 That the movement existed previously in some earlier form without the florid

 

2Rust, BG XVII, xv.
aspitta, Bach 11, 448, fn. 492.

155
vocal solo is thus quite likely. This hypothesis is supported further by the observation

that the cantata movements were probably copied from a model, as betrayed by the bar
lines in the autograph. In many instances the more florid organ discant spills over the
bar lines, the measures having been ruled for the ripieno which was copied in first. A
similar cramping of the keyboard parts in P 234 has already been noted earlier, and
again attests to the order in which the parts were copied. Also the oboe parts in the first
movement of the cantata are not obligato, but double string parts almost exclusively,
suggesting that they also were added later.4 Thus it seems tenable that both the concerto
and cantata go back to a lost model. At many points the concerto is more embellished
than the cantata, with many of these embellishments the result of corrections and
additions which are visible in P 234. Some passages in the cantata, however. are more
florid than in the concerto, suggesting that these may be closer to the original than the
reading in the concerto.

The Concerto in D Minor (1052) presents the most complex array of sources and
versions of any of the seven concertos in P 234. No fewer than four separate versions
of this work exist:

1) The autograph P 234.

2) A parts copy from the estate of CPE Bach, St 125. The cembalo part is in
Emanuel’s hand. Rust accepted it as autograph, and adopted its reading for the BG,
even though the cembalo part is at variance with P 234 at many points.

3) Parts copy St 350 preserves yet another version, commonly referred to as BWV

1052a. While the string parts differ only insignificantly from P 234, the solo

 

4Siegele, Kompositionsweise, 136.

156

cembalo part in St 350 is considerably simpler, with the left hand mostly doubling
the continua. This version is the most controversial, its authorship having been
disputed by many scholars.“

4) A parallel version to the first two movements is found in Cantata 146, while the
Sinfonia to Cantata 188 parallels the finale. The concerting instrument in both
cantatas is the organ, with the ripieno augmented by three oboes in the Sinfonia to
Cantata 146.

The extraordinary number and variety of versions of this work has encouraged
considerable discourse among scholars, particularly concerning questions of origins and
interdependence.“ Rust hypothesized that 1052 went back to a violin concerto, as
evidenced by the range, which stays above the violin’s lowest pitch, and by the
prevalence of idiomatic violinistic effects such as open strings and arpeggios. Spitta
further suggested that 1052a was a first, less artful arrangement of the work,
subsequently recast for Cantata 146, and finally arranged with greater polyphonic
development into 1052. Aber’s study (1913) showed that the corrections in the cembalo
bass and octave placements of the discant in P 234 supported Rust’s view and proposed
a now lost violin concerto as model. He also accepted Sebastian’s authorship for P 234

but posited Emmanuel as a likely author of St 350 (1052a), and further conjectured that

 

“Reproduced in the appendix to BC XVII, pp. 275-313. For the second movement
(pp. 291-297), all four versions of the cembalo discant are presented on adjacent staves:
”Erste Lesart" is 1052a, ”Zweite Lesart" is the organ discant of Cantata 146, "Dritte
Lesart" is P 234, and ”Vierte Lesart" is St 125.

“A summary of much of this discussion may be found in Siegele, Kompositionsweise,
102-103.

157
perhaps Vivaldi had authored the original violin concerto model.7 Paul Hirsch (1929)

held P 234 and St 350 to be independent of each other, one prepared by Sebastian, the
other by Emmanuel, and concluded that both versions probably went back to a violin
concerto model which, he hypothesized, was not the true original but an arrangement of
a still earlier concerto for a seven-string instrument, possibly viola d’amore.8 This
original model presumably derived not from Sebastian or Vivaldi. Hirsch also held that
for Cantata 146, only P 234 could have served as model. Siegele’s more recent careful
and extensive study of 1052, 1052a. and 146 has revealed that each shows developments
and modifications unique to itself.9 Noteworthy is the observation that 146 sometimes
more closely parallels 1052 while at other times is closer to 1052a. The obvious
conclusion is that none served as model for either or both of the others, but rather that
all three versions are independent and go back to a common, lost model.

Subsequent to Siegele’s study, Andreas Glockner made the remarkable
observation, unnoticed previously, that two compositional layers are perceivable in parts
copy St 350, confirmed by the presence of two different watermarks. The string parts
were copied first around 1734, but the cembalo part was completed only sometime later,
still before the appearance of P 234. This discovery led Glockner to offer two
hypotheses: 1) that Emmanuel attempted a first transcription of the violin concerto model

in 1734, but the cembalo part somehow was subsequently lost and survives only in a

 

7Aber, "Studien," 13 ff.

“Paul Hirsch, "Uber die Vorlage zum Klavierkonzert in d-moll,” BI 26 (1929): 153-
174. Hirsch’s hypothesis of viola d’amore was based on considerations of the playing
technique required and octave placements.

9Siegele, Kompositionsweise, 101-115.

158

later COPY; or 2) that the string parts of St 350 represent the original performance
material of the D Minor violin concerto model, the solo violin part of which is now lost.
At a later time, Emmanuel then undertook a transcription of the violin solo part for
cembalo.10 While neither of these hypotheses are provable, the second would appear
to have greater merit. The first hypothesis’ assumption that the solo part of the first
transcription was lost allows the possibility that the solo instrument in the 1734
arrangement may not even have been the cembalo. Equally add is the prospect of a
second copy of the presumed original cembalo solo prepared at a significantly later
time,11 since making such a copy obviously would require the use of the original to
produce the copy; i.e. the original was still extant when the copy was prepared. As
multiple copies of parts have not survived for any Bach concerto, the question arises why
a duplicate copy would have been made when the original was still available, particularly
if this copy had no further changes or alterations in it.12

While comparisons between the three versions reveal significant differences in the
keyboard parts, very few dissimilarities are found in the ripieno parts. In those instances
where differences do occur in the ripieno, the reading in 1052a (St 350) generally
appears to be the earlier and more formative one, while the reading in Cantata 146

implies that alterations and doublings were added. For example, at measure 93 in the

 

loGlc’ickner, ”Neuerkenntnisse," 55-56.

“Although 1734 to sometime prior to c. 1740 (the time of P 234) may not seem an
inordinate time span, Glockner considers it a ”not insignificant” gap. Ibid., 55.

12The only condition under which this copying process would make any possible
sense is if the original was lost and the copy of the cembalo part was retracted from the
score. However, it is also likely that the harpsichordist often played these concertos
from the score.

159

first movement the first violin part is doubled by second violin and viola in the cantata,
while no such doubling occurs in 1052a (Ex. 4-1). Since the cantata also makes use of ‘
a fuller instrumentation with three added oboe parts, the string doublings in the cantata
no doubt were added to help balance the richer sonority of strings with winds. Much

as the oboes represent an addition, so does the doubling of string parts.

Ex. 4-1. 1052a/146, I 93

1052a 146

 

Similarly at mm. 162ff. of the same movement the strings reenter after a long solo
episode with a light chordal accompaniment which is altogether absent in 1052a. Rather
inconsistently, 1052 follows 1052a at measure 93 but adopts the cantata reading at mm.
162ff. As with several other concertos without extant models, Wilfried Fischer has

attempted to reconstruct the original (lost) concerto which served as model for 1052.‘3

 

13NBA VII/7, 3-30.

160

It is noteworthy that Fischer’s reconstruction for this concerto follows the reading of
1052a for both of the passages just cited, since 1052a most likely represents the reading
closest to the original source.

Of similar interest is a comparison of variant readings of mm. 56-57 (Ex. 4-2).

Ex. 4-2. 1052a I, 56-57

      

K=__=_===-:__,__‘..__
____:.=__=!- ====

A-fi

 

A
- __ . ...._-.- _
=— 1____ = =.=__-—==_—- ~==E

mo

Elli

lath

III.

. UK.»

on'

1052, I 56-57

 

Comparing 1052a and Cantata 146 reveals that in the latter the passage at m. 56 is
realized via an exchange of parts: the viola part is derived from the continua bass of
1052a, while the continua part now doubles the organ bass, this bass now being the same
as the cembalo bass of 1052a and in all likelihood the original continua bass of the lost
model. Returning the pedal tone A’s to the continua part in the reconstruction is
certainly correct. but the viola might more rightly have been left as it appeared in 1052a
rather than following the reading in the cantata, since the latter is probably closest to the

original reading. The later 1052 version parallels the cantata reading rather than 1052a,

162

except for the cembalo bass, which follows the continua part from 1052a rather than the
static repeated pedal tone in the cantata. Such examples illustrate the complex of
relationships present between the variant versions of 1052, and support Siegele’s
conclusion that no single version is a direct descendent of another but that all go back
to a common (lost) model. Stylistic considerations nonetheless suggest that among the
three versions 10523 is the earliest, Cantata 146 stands second, and 1052 is the latest
rendering of the work.

Perhaps the most puzzling aspect of the three versions of this concerto is the
different octave placements of the keyboard parts. Generally, the cembalo discant in
1052a appears an octave higher than in the cantata, except in mm. 28-35, 140-143, 152-
156, and 179, where the discants of both versions are in the same octave. In all of these
cited passages, however, it is evident that the discant in 1052a is lowered an octave to

9

avoid e” and above, these pitches being unavailable on Bach’s harpsichord.”
Comparison of variant readings of mm. 152-156 illustrates this observation quite plainly

(Ex. 4-3).

 

1“As noted in chapter 111, the top limit of Bach’s harpsichord was d”’, and in none
of the harpsichord concertos is this pitch exceeded.

163
Ex. 4-3. 146/1052a1, 152-155

146 organ

    

 

1052a cembalo

 

Left unaltered. the cembalo discant in 1052a would rise to a’”, as in fact occurs in the

solo violin part of Fischer’s reconstructed violin concerto (Ex. 4-4).

Ex. 4-4. 10521,152-155 reconstruction

 

Consistent avoidance of pitches above d”’ on the keyboard by this rather prosaic process
has been seen by some scholars as betraying a less experienced hand. further supporting
the hypothesis that the arranger of 1052a was not Sebastian but a younger, less
experienced figure such as Emanuel.” Although there are extended episodes where the

discants of 1052 and the cantata are in the same octave,” the discant in the cantata

 

l"See Breig, "Entstehung,” 40-41.
l"Mm. 22—38 and 136-184.

164

generally lies an octave lower than in either 1052 or 1052a. extending downward even

to G (Ex. 4-5).

Ex. 4-5. 1461, 7-8 organ

 

Siegele and others have speculated that this unusually low range represents a "notational
peculiarity," and that Bach must have intended the concerting part on the organ to use
4’ registration.‘7 That the organ discant at 4’ pitch will sound at the unison with

ripieno violins in tutti sections, as at I, 13 ff.. does support this premise (Ex. 4-6).

Ex. 4-6. 1461, 13-15

 

 

l7Siegele, 105. Hermann Keller and Werner Breig also shared this view. Breig,
”Enstehung,” 40. See also NBA KB I/11.2, 83-85.

165

Also supporting 4’ registration are a number of instances where parallel fifths or octaves
occur between the organ discant and upper ripieno parts, as at mm. 74-75 of the first
movement.” If the organ discant sounds at the octave above, however, then the

parallel fifths invert to fourths (Ex. 4-7).

Ex. 4-7. Cantata 146 I, 74-75

- _ ..
v
- r
——_

 

Still more curious is the fact that the autograph of the concerto betrays signs of octave
modifications made by the composer, i.e., Bach in some passages apparently made

deliberate choices relative to the octave dispositions. In several instances, Bach at first

 

”The presence of such parallel intervals has encouraged some scholars to question
the authenticity of this work. Johannes Schreyer, Arnold Schering, and Paul Brainard
all have questioned the genuineness of the work, while Siegele, Fischer, and Breig have
argued for its legitimacy. In either case, there remains the question of the authorship
of the source concerto. Reinmar Emans in the NBA concludes that Bach’s authorship
will remain assumed until evidence to the contrary is brought forward. NBA KB l/11.2,
82-83.

166

copied in one reading only to cross it out and raise or lower it an octave, as at mm. 136-
138 of the first movement (Ex. 4-8) where the cembalo discant was first written in the I
upper staff, subsequently crossed out, and rewritten an octave lower in the bass staff.

The higher reading is found in 1052a, the lower reading in Cantata 146.

Ex. 4-8. 1052 I. 136-138 P 234

 

 

'II: III Pz| ’3

_. mi
. -.

 

Similarly in the second movement at mm. 50-52, Bach first wrote in the left hand pitches
beginning on c“, then duplicated the note heads an octave lower on the same stems,
finally adding the directive "octav tiefer" over the passage (Ex. 4-9). In this instance,

both the cantata and 1052a have the higher reading.

Ex. 4—9. 1052 II. 50—52 P 234

 

167

No definitive pattern seems to emerge from these octave disposition puzzles aside from
the general conclusion that the discant of 1052a is set high except in those passages
where d’ ’ ’ would be exceeded, and that Cantata 146 is generally notated an octave below
1052a. The very low tessitura in the cantata as notated results in the organ discant
occasionally crossing below the organ bass and continua, a very curious phenomenon
which sometimes results in second inversions in the harmony (unless the part sounds an
octave higher).19 For 1052, Bach generally followed the upper octave range used in
1052a except where 1052a artificially accomodated the top pitch limit of the harpsichord.
By dropping an entire passage rather than just a few problematic pitches to the same
lower octave as the cantata, the original voice leading of the lost model is more
accurately preserved. The difference is evident by comparing mm. 140 to 146 of the

first movement (Ex. 4-10).

 

19Such part crossings may be found in the first movement at mm. 14, 15, 17, 45-46,
101, 106, and 107, and in the second movement at mm. 27, 53-55, 69-71, and 75.

168
Ex. 4-10. 10521, 140-146

Reconstruction violin

 

 

 

 

 

 

1052a cembalo

 

1052 cembalo

 

Not only has Bach successfully accomodated the extraordinarily high violin tessitura to
the more limited gamut of the harpsichord and preserved the proper voice leading. but
at the same time has expanded the actual range of the harpsichord discant, despite the
upper d’” limit. Comparison of the range of the discant in the variant versions is

instructive:

169

Version Disgnt Range
1052a g - d”’ (2 octaves + 5th)
146 G - a” (3 octaves + 2nd)
1052 G” - d”’ (3 octaves + 5th)
Vl. Concerto g - a”’ (3 octaves + 2nd)

(NBA Reconstruction)

In essence, the original range is found in 146 (notated down an octave) and in the
reconstruction, while 1052a and 1052 represent adjustments and accomodations; 1052a
shrinks the overall range, while 1052 expands it. Particularly noteworthy is the
observation that Bach’s approach to accomodating the range of the original violin part
is quite different from that encountered in other concertos in P 234. In the other six
concertos, the typical cembalo discant range is f (g transposed down a step) to d”’.
Thus the original violin range is accomodated without overstepping the bottom limit of
the violin’s compass (g; f in transposition), while the upper limit of the harpsichord
(d”’) is accommodated by brief alterations of voice leading and octave transpositions.
In none of them, however, does one find entire extended sections set an octave lower as
a means of remaining below d”’, as appears to be the case in 1052 I, 136-184. Siegele
concluded that the cantata had the original voice leading and octave placements, if one
accepts the 4’ registration. The foregoing observations support his judgement, although
the question of 4’ registration remains open.

Siegele further noted that the three unison tuttis in the opening movement (mm.
1-7, 172-173, and 184-190) are treated differently in the variant versions. In the cantata,

the organ discant drops out during all three tutti episodes, while in 1052 a right hand

170

part is supplied which doubles the violins. In 1052a the cembalo discant is silent in the
opening tutti (mm. 1-7). but doubles the ripieno parts at mm. 172-173 and 184-190,
much as in 1052. The question which thus arises is whether or not the solo instrument
played during the tuttis in the original version. Without giving particular reasons,
Siegele holds that the cantata represents the probable reading of the lost model in all
three cases, i.e., without doublings. This is problematic, however. Both in this cantata
and in the other obbligato organ cantatas which parallel these concertos, the realized
discant drops out in tuttis, especially during the opening ritornello (so that the organ may
furnish continua harmony). By contrast, in the extant violin concertos (1041-1043) the
solo violin always doubles the ripieno violins in the opening tuttis.” Since the
presumed model for 1052/1052a is a violin concerto, it is very likely that the solo
instrument--the violin--did indeed double the ripieno violins. Appropriately, then,
Fischer’s reconstructed violin concerto in the NBA follows this pattern, with the solo
violin doubling the ripieno strings in all three unison tutti episodes. When adapted for
use in the cantata, the organ discant, which was fashioned from the solo violin part of
the model, was dropped during the tuttis in keeping with the common practice in
cantatas.

The absence of a parallel cantata version to the third movement of 1052 (the lost
Sinfonia to Cantata 188) makes definitive conclusions about the original model concerto

somewhat more tenuous, but close scrutiny of the autograph reveals that at many points

 

”In the A minor Concerto for Harpsichord, Flute, Violin, and Strings, BWV 1044,
the cembalo discant falls silent during the opening tutti, but the solo violin doubles the
ripieno violins. Wherever the cembalo discant drops out during tuttis in this concerto,
the bass line is figured, suggesting that the harpsichord functions as a continua rather
than solo instrument.

171

the reading in P 234 prior to corrections is in agreement with the reading in St 350
(1052a).” Of particular note is the solo passage at mm. 73-77 where the autograph is ‘
heavily reworked and three stages of compositional activity are discernible. Bach first
copied in the sustained d’ found also in 1052a, a reading which, while suitable for violin.
is totally unsatisfactory on the harpsichord due to its rapid sound decay. He then
modified this reading by substituting a doubling of the viola part, and finally arrived at

the richer figuration in continuous sixteenths (Ex. 4-11).

Ex. 4-11. 1052 III, 73—77

P 234

 

 

21A list of these before-correction readings is given in Siegele, Kompositionsweise,
107, ms.

172

Cembalo discant, first version (=1052a)

 

Cembalo discant, second version

  

Cembalo discant, final reading

 

In the parallel passage at mm. 177-180, the first reading doubles the viola and is
subsequently enriched with sixteenth figuration paralleling the reading at mm. 73-77.
The agreement of P 234 before corrections with St 350 gives reasonable assurance that
the original version generally matched St 350; the reconstructed violin concerto in the
NBA adopts this reading.

Similar observations and conclusions may be drawn from the Concerto in E
major, 1053, the autograph of which again betrays many corrections and alterations.
The readings before the corrections agree substantially with the readings in Cantatas 169
and 49, the parallel versions to this concerto, and lead to the conclusion that both
concerto and cantata go back to an earlier form. Buttressing this hypothesis is Siegele’s
observation that the width of the measures in the cantata autograph are often too narrow
to accomodate the organ part, much as was the case in cantata 146, indicating that the

ripieno parts were copied first from another source. Also noteworthy is the fact that the

173

three oboe parts in the Sinfonia to Cantata 169 are not obbligato, but virtually always
double string parts to add color, suggesting that they also were added later. In the few
instances where oboes play by themselves without ripieno strings, there can be little
doubt that such passages were originally set for first and second violins, as is done in
1053 I.”

The principal if not unanswerable questions about 1053 concern the key and solo
instrument of the model, as there are far fewer clues to help determine the original form
of this concerto than was true for 1052 with its multiple versions. Spitta and Schweitzer
thought the original model was a concerto for clavier,23 but other instruments have been
suggested since, including flute, oboe, or violin. The question of instrument is linked
in part to a determination of the original key. Comparison of keys between the concerto

and parallel cantata movements shows the following:

Mov. Concerto Key Cantata Key
I E 169, I D
II c” 169, V b
III E 49, I E

In the concertos with extant models, it was noted that the key transposition was
uniformly downward to accomodate the harpsichord’s more limited range. That the keys
of the first two movements of 1053 are higher than in the cantata thus raises speculation
whether the cantata key is necessarily the same as that of the original model. Fischer

noted a few occasions in the first movement (mm. 13, 99, and 122) where the second

 

”Brief passages of oboes without strings occur at mm. 13-14, 23-24, 66, and 71-72.

”Spitta, Bach, III, 137; Albert Schweitzer, J. S. Bach (London: A. & C. Black,
1935), I, 411.

174

violin plays g”, which would thus require f" (if available) in the cantata. In each
instance, the pitch is altered in the cantata to accomodate the range of the violin,
suggesting that the original key was higher--perhaps E“, E, or even F. Siegele proposed
two possibilities: an oboe concerto in E“, or a flute concerto in F, with the first more
likely.” Fischer’s further observation of several instances in the autograph where Bach
first wrote in a single sharp accidental, subsequently corrected to a double sharp, also
supports Siegele’s E“ hypothesis, since a single sharp is redundant in E major, but would
make sense in either Eb or F major.” It is apparent that many questions remain, so that
determination of the original form and key of 1053 remains uncertain. Given the
differences in keys and scoring between the movements, it might be questioned whether
the three movements of the concerto even stem from the same model.

The following table lists known or presumed keys for the various versions of the

harpsichord concertos (Table 4-1). Keys in brackets are conjectural.”

 

2“Siegele, Kompositionsweise, 142-143.
”Wilfried Fischer, NBA KB VII 7, 137.

”Conjectured keys in brackets reflect hypotheses in current Bach scholarship,
principally from Siegele and the Berichte of the NBA.

\v.q¢qn\wmutmi.wlnw

175
Canceng $9an Key Orig. key Parallel Version Key

1052 d [d]27 Cantata 146 I, (I
II g Cantata 146 V, g
111 d Cantata 188 I, ?

1053 E [E“, F] Cantata 1691, D
II c” Cantata 169 V, b
111 E Cantata 49 I, E

1054 D E

1055 A [A]

1056 f [g]
II A“ [B“] Cantata 156 I, F

1057 F G

1058 g a

1059 d [d] Cantata 35 I, d”

1060 c c

1062 c (I

Table 4-1. Keys of Concertos

As based on known or likely keys of the models, it becomes evident that the concerto
transcriptions virtually always are transposed down, the only exception being 1053 if the
original model was in E“. It is also possible of course that the highest pitch in the model

for 1053 may not have necessitated lowering the key to accomodate the cembalo; in fact,

 

27Given the range of the solo part in this concerto (lowest pitch the open g string),
D minor appears the most likely key. However, there is a brief passage in the autograph
(I, 9 4.beat through 12 1.beat) where the cembalo left hand part is notated incorrectly
a step too high and is rather obviously corrected via thickened note heads in mm. 11-12,
suggesting that the score may have been copied from a source in E minor. It is difficult
to speculate on another cause. The range of the music in all other respects supports D
minor, however.

”Rust noted in his preface to the cantata (BG 7, xxvii) that cantata movements I and
V were in a neater hand than the remaining movements, which, when considered
alongside the existence of the fragmentary Concerto in D Minor, 1059, led him to the
conclusion that these movements went back to an earlier instrumental concerto. Spitta,
Siegele, Fischer, and Rifltin all reconsidered the question subsequently and generally
concur in the existence of an earlier source concerto. probably for oboe. It is notable
that both Cantata 35 and the Concerto 1059 include oboes in the scoring. NBA KB I/20,
187-8, 193-4; KB VII/7, 138-140.

176

there is but one instance where the cembalo part appears to be altered to avoid

overreaching d”’ (Ex. 4-12).

Ex. 4-12. Cantata 1691, 110-112 organ

 

 

(lam.

10531, 111-113 cembalo

 

In this example, the rising melodic line clearly is lowered an octave to avoid d””’.
The question of key transposition is still more clouded when turning to the second
movement of 1056, which has a parallel in the Sinfonia to Cantata 156. Since the time
of Rust, a (lost) violin concerto in G Minor has been generally acknowledged as the
likely model for this concerto. For a variety of reasons, however, it seems unlikely that
the present second movement belonged to this violin concerto originally.” Since the
second movement of 1056 is in A“, the corresponding movement in the violin concerto

bass

would have been in B“, with the range of the solo violin g’ to e , unlikely inasmuch
as the bottom octave of the violin is completely unused. Further, the bottom of the

cembalo range in the outer movements is C, but in the middle movement is the very

 

”For a fuller discussion of questions related to the second movement and its
provenance, see NBA KB VII/7, 81-86.

177

bottom of the harpsichord gamut, low GG. Perhaps most telling is the different handling
of the bass: while in the outer movements the cembalo left hand either doubles the
continua bass or carries the bass during solo episodes when the continua part drops out,
in the middle movement the bass is carried by the continua, the cembalo playing a lightly
embellished version. These observations suggest that the middle movement may not
have belonged with the outer two originally, and hence both the source and key for the
second movement remain unknowable.’0

Since both 1052 and 1053 have parallel versions in cantatas with obbligato organ,
comparisons become inevitable. Unlike Cantata 146, the organ discants in Cantatas 169
and 49 are notated with the same octave placement as the cembalo in the parallel
concerto, thus obviating any hypothesis about 4’ organ registration. There is no apparent
reason for this difference in octave placement. That the organ discant is at 8’ pitch in
Cantatas 169 and 49 suggests that it is Cantata 146 which has the exceptional octave
placement.

It may also be worth noting that cantatas 169 and 49 (paralleled in 1053)
originally appeared within two weeks of each other in the fall of 1726, several months
after Bach’s failure to obtain a favorable settlement of grievances against the University

from King Frederick Augustus. The king’s response to Bach’s complaint is dated

 

”Joshua Rifkin has offered a most interesting hypothesis, suggesting that the Sinfonia
to Cantata 156 and the middle movement of 1056 both go back to a (lost) oboe concerto,
the outer movements of which are preserved in the first and fifth movements of Cantata
35 (the first movement of the latter paralleled in the fragmentary 1059). NBA KB I/20,
194, and Joshua Riflcin, "Ein langsamer Konzertsatz Johann Sebastian Bachs,” B] 64
(1978): 140-147.

C1,...“ mm

178
Dresden, January 21, 1726.31 Cantata 72, performed the following Sunday on January

27th, was the last of Bach’s own cantatas offered to the Leipzig congregations from early
February through May. In these months he substituted lesser works by his Meiningen
cousin Johann Ludwig. While Bach had written cantatas with obbligato organ
previously, nowhere is their appearance as concentrated as in the fall of 1726. Cantata
169 was performed on October 20, and Cantata 49 on November 3. In addition, there
are three more cantatas with obbligato organ in close proximity between September 28
and November 3, including Cantata 35 which parallels the fragmentary Concerto in D
Minor, BWV 1059.” One might speculate that the cantatas with organ obbligato, while
fresh works from Sebastian, nonetheless required a less concerted effort and expenditure
of time, given that the principal Sinfonia movements were reworkings of already existent
earlier concertos, and the balance comprised a solo cantata. If, as suggested earlier, the
first two movements and last movement of 1053 stem originally from two different
source concertos. then perhaps the chronological proximity of the parallel cantata
versions suggested bringing these concerto movements together into one concerto
transcription for harpsichord. The cantatas which are paralleled in 1052 (Cantata 146

and 188) are datable with less surety, but may stem from April and October of 1728.

 

3‘The Bach Reader, 105.

”The other cantatas from the period are Cantata 27 from October 6, and Cantata 47
from one week later, October 13. Cantata 35 was performed on September 28. Datings
from Christoph Wolff et al., The New Grove Bach Family (New York: W. W. Norton
and Company, 1983), 178-189, which in turn are based on the chronology of Bach’s
vocal works established by Alfred Diirr and Georg von Dadelsen.

179

Fashioning the Cembalo Parts

As with the concertos examined earlier, the extent of reworking of cembalo parts
is again of primary interest. For the concertos with extant models (1054, 1057, 1058,
1062). the availability of the source works permitted direct and definitive comparison.
It should be obvious, then, that the same degree of conclusiveness cannot be expected
or presumed when examining concertos for which extant models are lacking. The
existence of parallel versions in the present group of concertos still allows reasonable
conclusions to be drawn, however, provided it is recognized that the parallel version may
itself be a reworking of the original.

When examining the group of concertos with extant models, it was noted that
Bach’s adaptations often sought to accomodate violinistic figuration to the keyboard and
at the same time make the writing more brilliant. While 1052 and 1056 in all likelihood
were violin concertos originally, the parallel versions which survive are themselves
arrangements for keyboard (obbligato organ in the cantatas, harpsichord in 1052a), so
that the adaptation of string idioms to keyboard in the concertos is far less apparent.
This is especially true of 1052, where the cembalo discant is virtually identical to the
organ treble of Cantata 146 and the right hand part of the earlier 1052a version
preserved in manuscript St 350. One of the few tell-tale vestiges of the original violin
writing may still be seen, however, in the widely spaced intervals so idiomatic to violin
writing which are yet present at m. 27 of the first movement. In Ex. 4-13, note that the
awkward skips, widening to two octaves, are preserved in the cantata, avoided altogether

in 1052a, accomodated by a division of hands in 1052 (with chords added to increase

mu

180

sonority and the sense of climax), and restored to their "natural" state in the violin

concerto reconstruction in the NBA.

Ex. 4-13. 146/1052 I, 26—28 keyboard parts

Cantata I46

_.‘—_—
n
I'..——_.__--—-- —— I
‘-_—-='—:r---r— L'—=E---_-:- :---rl=--------_-_-------=
, < _ a . ,

 

1052 reconstruction

-————— —-— v _—
_—-———-—--v _-' - r! —-- _‘u—'=- -- u- l- - n- n.—
_——-r-_-r— —v-— _- -=--=----—-
p '- _. __ .— _

 

It becomes apparent that Bach’s adaptation of this passage in 1052 takes not only the

musical effect into account, but also considers the technical practicalities.

181

The E major Concerto, 1053, presents many more examples of embellishment
than are found in 1052, particularly in the discant. The extent of this ornamentation is
evident already in the first solo episode, where scalar flourishes in smaller rhythmic
values, written out turns (m. 14), and even the insertion of an entire new measure (m.

13 in the concerto) all exploit the capabilities of a keyboard instrument (Ex. 4-14).

Ex. 4-14. 169/1053 I, 12-15

169 organ

 

1053 cembalo

 

Elaboration of the treble in the cembalo discant is perhaps nowhere better illustrated,
however, than in the second movement of 1056, a richly ornamented cantilena played
not by violin but by oboe in the parallel Sinfonia to Cantata 156. While the cembalo
bass is comprised of a fairly simple eighth-note pattern derived from the continua part,
the discant is afforded particularly lavish treatment, so much so that Otto Kinkeldy
regarded it as a notable example of Bach’s ornamentation - practice.” The
embellishments clearly are worked into the autograph as a result of corrections. and the

reading before corrections, as far as can be discerned, is largely in agreement with the

 

”Siegele, 129.

182

reading in the cantata.” The extent of the ornamentation applied to the discant in this

movement is almost without equal in these concertos, and is illustrated in Ex. 4-15.

Ex. 4-15. 156/1056 II, 9-12

156 oboe

 

1056 cembalo (up a 3rd)

 

Derivation of the Cembalo Bass

Of greater import than the embellishment of the discant, however, is the
reworking of the cembalo bass, which again is derived to a lesser or greater degree from
the continua bass of the model. As noted earlier, it is this aspect of these concertos
which largely establishes the solo character of the cembalo and the validity of the genre.
Although original versions of the concertos in this group are lacking, it is not
unreasonable to assume that the continua bass as it appears in the parallel versions is
reasonably faithful to the original, particularly when corroborated by agreement with the

continua found in the concerto. Thus comparisons between the cembalo bass, continua,

 

34Essentially. the discant before changes matches the oboe part of the cantata with
only a few slight differences. The reading of the autograph before corrections was
published by Aber in ”Studien,” 27-29.

183

and continua from the original (so far as it is assumed to be represented in the parallel
versions) again may be considered, as well as the addition of chordal "filler. "

To a considerable extent, the amount of agreement between the bass parts not
only corroborates the likely original continua bass, but also distinguishes the level of
compositional mastery exhibited in the transcribing process itself. This point is best
demonstrated in the D Minor Concerto (1052), which has the advantage of three separate
versions (1052, 1052a, 146) to authenticate the bass. The continua parts of 1052 and
146 are for all intents identical;” the continua bass of 1052a is also in agreement,
excepting a few minor exhanges of parts, as at mm. 28-33 where the repeated eighth-
note rhythm found in the continua parts of both 1052 and 146 is taken up instead by the
cembalo bass while the continua is replaced with the simpler quarter-note quarter-rest

pattern used for the organ bass in 146 (Ex. 4-16).

 

35Only very minor differences appear in mm. 59-60, 83-90, and 162-172. In the
latter two cases, the only difference concerns the length of time for which the continua
part drops out in one version or the other; there is no net difference in harmonic content.

184
Ex. 4-16. 146/1052/1052a I, 28-32

146 organ and cant.

         
 

ml"

Pb=—————
\"rO-'l-O'|‘.‘IOI‘ raw". ‘
_—--——ll__l—‘ ‘

 

1052a cembalo and cant.

 

_____
_-__..__ ===.-_—==— —
. _-—_ _—-
‘1 piano
1 I‘
1 _._. __ ._u_.-__..__ _._.—_._...—
Ear. .—..-=..—..'=:-;.-:=.-===:=—_ .===-;_-:==:—_-=§_=:=‘=—1r5§_.—r;-=E=_=:==L‘£E_
.- ‘1 .- -—"‘ \‘ '-- -- -- II- '=‘=_-=----‘~==
2.. ___ ___ __- y” ___ -__ \
_ _ _ m
E $=;;=‘=§====e=====-====:===~=5=.==.====

There is no apparent purpose served by this exchange of parts other than to give the
cembalo bass a stronger rhythmic element (as compared to the cantata). The repeated
eighth-note figure appears a number of times in the continua bass throughout this
movement, each time doubled by the cembalo bass. Although the left hand part of 1052

also parallels the continua, the constant octave leaps add interest, particularly when

185

compared with 10523. One of the conspicuous features of first movement of the latter
is the rather rigid doubling of the continua by the cembalo bass; only when both hands
are required to negotiate the figuration does the cembalo bass free itself from the
continua (mm. 62-68, 71-75, 95-104, and 146—153). This characteristic is also found
in the second and third movements; in a few solo episodes of the latter, the continua part
drops out entirely while the cembalo bass takes over the continua bass part. In the
second movement, the lengthy solo is handled the same way: the cembalo bass assumes
the role of continua, the latter remaining silent except for opening and closing
ritornellos. The end result of all this is a somewhat mechanically generated transcription
in which neither left nor right hand parts engage in any independent exploration of
embellishment or thematic manipulation. Furthermore, there is no chordal filler added
into the texture. Thus 1052a represents the simplest type of concerto transcription as
identified earlier: the cembalo bass is taken directly from the original continua part; the
new continua part simply doubles or plays a simplified version of this bass; and little or
no chordal filler is employed.

By contrast, the more familiar version of this concerto from P 234 (1052), despite
a very limited use of embellishment in the discant, treats the cembalo bass very
differently. Comparison of the cembalo bass part among the various versions in the first

solo episode makes this difference immediately apparent (Ex. 4-17).

186
Ex. 4—17. 1052 1, 7—13

1052a cembalo and continua

----- n.‘ .-—‘..- .-“--—--—=-— .‘n- m— --'I-
' ===_ _. _- =__.=_ -= _.‘—J- n-4--—‘_—_-.—_-

=- - :‘ -‘- - - --=- ---l_---- -“-- l-- -— -=
==- - H =‘_' —--— --—:— -—-—-— —- —

-—====—————==.:—-
#

 

 

  

-
_.‘

tux—.v .._-—- -—_—===- -—— _— - -‘ _-_. -’-T__-;—-i-

nan—v n‘-— -n --———'- _.’..‘_ -.-————.y—-——— —_— -.-—————

u--—..-—————.-- -.-—_—-_ n-‘—_———__—_—_—_ _-__-=_ _-_;_-__

_-—.:— _._.. — u- _— ——_- 7H_,____.2,,,__'_ _— =

    

   
 

 

-u

    
 

———-————=—_—-—7;——-~ —
_——_—-——.-—-——-——;:" ...-=—:
——-=— —- _--——I=-m———-

    

::_ .-

n-n-v—-——__— - ._-__—.__.-.

, .- -—.-————— —_:=——-——'——_

.,,: , .- ‘u- -_———=—--—-—- u. -—_.r- _—

-_ E—u—.—-- --» .=-.. L..- _—_ -_'-
—_

      

 

 

187

What stands out so conspicuously in the last reading is the wonderfully inventive and
brilliant left hand part which complements the discant but also far transcends the skeletal
tonic repetition in the continua. The entrance of the left hand at m. 8 imitates the solo
treble entrance of a bar earlier, producing a contrapuntal tonal answer to the discant
which is not even hinted at in the other readings. The entire movement is replete with
examples of a similar nature in which the left hand part is freed to explore all measure
of contrapuntal and figurative device. The following two examples illustrate what are
basically arpeggio and scalar patterns joined and fashioned into motivic sequences, yet

still dependent on the basic harmonic outline established by the continua (Ex. 4-18).

Ex. 4—18. 1052 I, 34-41 cembalo and continua

 

“_—
~

188
1052 I, 104-109 cembalo and continua

 

A further divergence from 1052a is accomplished by the addition of significant
amounts of chordal filler in 1052, found most frequently in the discant but also
encountered in the bass. It is particularly noteworthy that many of these added chord
tones are not simply doublings of ripieno string parts, but are fully independent (Ex. 4-

19).

Ex. 4—19. 10521, 17—21

”_—
~-__ —- _ 1—
r'|_ “Ln—$4

__—__n———_

4
.....-— ——-v Hm—

I' —n.. L...— —_--——

.-=—— ‘14-..— —-___-_—_ -———-—r

 

1052 I, 27—28

 

From the autograph it is apparent that some of these chord tones were worked into the
text after the initial copying of the score, thus representing a second compositional stage.
Evidence of such additions sometimes appears as lighter script or superimposed note
beams, as in the following. The added text is conspicuous when compared with the

reading from 1052a (Ex. 4—20).

Ex. 4-20. 1052a/1052 Cembalo

1052a I, 112-113

 

1052 I, 112—113 P 234

               

 

 

.——.. .. ..._'_..’;T.-~ ' '

~ ,.,,,_ 1;- £3:-
zggl ”Ll-{$529}:

"""""'~"-‘-" ‘- ' " -=3’::="'

 

 

.- -.—————‘-————— _' u u .
t t..‘ . -— . , v A 1
, .-.. ,7 r 1
~ . . . . . ., . . 1,- _._

 

1052a I, 146-148

  

 

 

1052 I, 146—148 P 234

' ' I

 

In the last example, it should be noted that the beaming of the sixteenth notes in the
discant changes at m. 148, since the original reading at this point included the double
thirds; when similar double notes were added into the previous two bars, the note heads
and beams were written in on top of the existing notation.

It is unfortunate that Rust did not follow P 234 more faithfully when preparing
the BG edition of this concerto, but instead used the later reading of St 125 which adds
many additional chord tones and octave doublings in the bass which are foreign to the
autograph.” The extent of difference between these two readings is nowhere more
pronounced than in mm. 133-140. The added chordal filler at the end of m. 133 in the

BG does not appear in the autograph, but in the following bars thirds and sixths which

 

3“Parts copy St 125, from two unknown copyists, contains more embellishments than
P 234. Whether these additional embellishments derive from Bach or the copyists is an
open question. Wilfried Fischer has suggested that there may have been a further
corrected parts copy between P 234 and St 125, and that St 125 is only indirectly
dependent on P 234. NBA KB VII/7, 41. Rust’s view was that St 125 represented
Bach’s last and final thoughts on this work, and accordingly was given preference.
Rust’s list of variant readings between the two sources given in his preface to the BG
edition. however, does not begin to fully represent the differences. BG 17, xvi.

191

double upper strings are clearly present, and the bass in mm. 134-135 is an octave

higher than in St 125, thus in the same octave as the continua (Ex. 4-21).

Ex. 4—21. 1052 I, 133-140

(BC over P 234, 133-6)

 

192

(86 over P 234, 137-140)

Ex. 4-21 (cont.)

                  

 

:-

 

‘iII'I—ll'flcu.‘ III p

L
I»

 

 

.

J
.1.
.

 

 

. _‘_

'59}.
'97:;

-’j'{_-'_n0:’._l
PW l

e on ‘ .lav-v‘
.

      

  

 

 

”Alum—V am»—
. ,
-

_._—
_—-——
_—
-— — —
—
- u--.-
- _—
_—
—~i .
- ———
————
‘
4 .4

. 52.. o e .
.._-- ".—. UL.‘ ..__—_-.-.
. .

Transcription of P 234 cembalo

 

Where OC‘
actually v
melodica
C
to the a1
that mu
for 10:
manne
custo
0%!
pass
aga
10?
he

th

193
Where octave doublings of the bass appear in the BG (mm. 136, 138, and 139), Bach

actually wrote fifths and thirds above the bass, pitches which do not double string parts
melodically but fill in the harmony.

Comparable study of the third movement of the D Minor Concerto is limited due
to the absence of the Sinfonia to Cantata 188, but comparison with 1052a corroborates
that much the same compositional procedure is present. To begin with, the continua part
for 1052a is virtually identical with the continua in 1052. There is a difference in the
manner in which this continua bass is employed, however. In 1052a the cembalo bass
customarily doubles the continua, except in some solo episodes where the bass is given
over to the cembalo left hand while the basso continua part drops out.37 If in such
passages the solo figuration requires the use of both hands, the continua part may reenter
again to carry the bass. In this regard, then, 1052a shows some similarities to 1054 and
1057. Further, there is no chordal filler in this movement in 1052a. By contrast,
however, there is considerable added chordal filler in both solo and tutti sections in the
third movement of 1052, and although the continua bass and cembalo bass mostly double
each other, there are sections where the continua carries the bass and thus frees the left
hand to pursue its own figuration. It is also significant that the continua part plays
throughout in this movement, i.e., the bass is never passed exclusively to the cembalo.
Thus the continua becomes the real bearer of the bass part, the cembalo left hand either

doubling it or pursuing free figuration (Ex. 4-22).

 

37Examples include mm. 18-22, 24-28, 92-112, 130-138, 150-160, 230-243.

194
Ex. 4-22. 1052 III, 46-55 continua and cembalo

 

In contrast to 1052a, then, 1052 shows features seen in the Concerto in F Major,

1057, or even the Concerto in G Minor, 1058. Characteristic elements: the cembalo
discant follows the original violin solo part fairly closely, opening the way to greater
brilliance and exploitation of the cembalo bass; the latter, though sometimes doubling
the continua, frequently is freed from the continua to assume greater independence and
explore novel figuration, especially in the first movement, while the continua part proper
is left to carry the bass; and a fair amount of chordal filler is added into both discant and
bass. Like 1057, there is the inclusion of significant amounts of chordal filler, absent
from 1058; but the brilliance of the left hand part is stronger than in 1057, and thus
more akin to 1058. Nonetheless, the added brilliance of the left hand part, coupled with
a significant use of chordal filler, suggests that the cembalo here combines to some
degree the roles of solo and continua instrument, but with the greater weight falling to
the solo role. The E major concerto, 1053. has certain affinities with 1052, but

also some differences. Comparison of the concerto with Cantata 169, its parallel

195

version, reveals significant variances between the continua bass in the cantata and the
continua bass in the concerto, variances which are not traceable to corrections in the
autograph. Since both cantata and concerto are based on a lost model, the unanswerable
question arises of which reading is more faithful to the original. Since the continua bass
part in the concerto is sufficiently close to the bass of the cantata, it is possible to
speculate what the continua bass part probably looked like originally, and thus to
determine its placement in the concerto.

As has been true in several of these concertos, the cembalo bass in 1053 is closest
to the continua in tutti sections, though there are passages even in tuttis where the left
hand part digresses significantly from the continua. In mm. 15-20 of the first
movement, the cembalo bass parallels the continua line only to a point; in mm. 17-19,
the left hand roughly follows a composite of the continua part of the concerto and the

continua part from the cantata (Ex. 4-23).

196

Ex. 4-23. 169/1053 I, 15-20

169, 14-19 organ

 

10! 7 H ““08“. 1

1053, 15-20” continua and cembalo

A.

 

In solo episodes, it becomes plain that the principal carrier of the bass is actually the
cembalo left hand part, since the continua part either doubles the cembalo bass,
simplifies it. or drops out altogether. This procedure ensures that the cembalo bass
always plays a part which is at least as active as the continua, and often ornamented to
such degree as to far exceed the continua, adding considerable brilliance to the solo
cembalo writing. That the continua drops out altogether for extended segments strongly

suggests the decreasing importance of the continua. It is also worth noting that in the

 

3“There is an additional measure (m. 13) in 1053 compared with 169; hence the
difference in measure numbers.

197

two longest passages where the continua drops out,” the ripieno strings are also silent,
leaving the solo cembalo completely by itself. Owing to the lack of other parts to fill
out the harmony, Bach adds chordal filler into these measures, underscoring the
harpsichord’s combined solo and continua functions in this context. It is also noteworthy
that the harmony thus added is an exact realization of the figured bass in the

corresponding measures of the cantata (Ex. 4-24).

Ex. 4-24. 169/1053 1, 9-14

169 1, 9-13 organ w/fig. bass

 

I I ‘
D

1053 1, 9-14 cembalo

  

 

By contrast, in other solo episodes where the continua drops out for shorter periods or

the ripieno string parts remain, there is no chordal filler.‘0

 

”Mm. 9-13 and 122-126, the latter a complete restatement of the same passage in
the da capo ritornello.

”See mm. 27-30, 37-40. 67-69, 77-79, 81-84.

198

The same pattern is followed in the second and third movments: in tuttis, the
cembalo bass part doubles or ornaments the continua part, while in solo episodes the '
continua frequently drops out and the cembalo left hand part carries the bass. The
declining importance of the continua is especially apparent in the second movement,
where the continua drops out altogether during the extended solo which comprises the
majority of the movement (mm. 7-30). This Siciliano has certain affinities with the
middle movement of 1056, which also features a richly embellished solo melody, light
chordal accompaniment in the strings, and a walking bass line in syncopated broken

octaves in the cembalo bass (Ex. 4—25).

Ex. 4—25. 1056 II, 8—11

——_ —
— — ———-—---_vv——
-—-—- —--* ---_—

_._.—_—

 

199

The principal--and significant--difference lies in the function of the cembalo bass,
however. Whereas in 1056 the original continua is preserved in the concerto continua
while the cembalo bass has a modest variant of it, in 1053 the borrowed bass is in the
cembalo bass during the solo episode, the continua having dropped out. This distribution
of parts further corroborates the observation made above that for 1053, the cembalo left
hand part rather than the continua is the principal carrier of the bass.

Some chordal filler is to be found in 1053, particularly in certain solo
episodes.‘1 The combination of chordal filler plus the dependence on the left hand for
the bass parallels procedures encountered in 1054 and 1057, where the cembalo performs
the dual role of solo and continua instrument. This pattern is somewhat true in 1052
also, except that 1052 explores a much more brilliant keyboard technique, particularly
in the left hand part which often exercises considerable freedom from the continua part.

Thus the basic approaches to transcription found in the concertos with extant
models is to be found also in these concertos with parallel versions in cantatas. In the
simplest terms, two approaches may be distinguished: concertos in which the cembalo
still betrays affinity with traditional continua practice, reflected in at least some use of
chordal filler and a reliance on the left hand part to carry the original bass; and concertos
in which the left hand is often freed from the obligation to supply the original bass and
chordal filler is either absent or used more sparingly. In some works, elements of both

types may be combined to a certain degree.

 

“In movement I, mm. 9-14, 42-44, 63-65, 78-79. 122-128, and 155-157. There are
no added chords in the second movement, while the third has relatively few examples,
e.g. mm. 3-4, 45-46, 81-82, etc., all parallel passages at the return of the ritornello,
except for mm. 239-242 where Bach adds an alto part to the right hand, thus enriching
the harmony.

V.

The Cembalo in Concertos without Extant Models

The preceding chapters have examined the concertos with extant models (Chapter
2 and 3) as well as concertos with parallel versions in certain cantatas (Chapter 4).
There are two remaining concertos in P 234 which together encompass five movements
without extant models or parallel versions:
Concerto in A Major, BWV 1055
I. Allegro
II. Larghetto

III. Allegro ma non tanto

Concerto in F Minor, BWV 1056

I. [Allegro]
II. Largo (Parallel version in Cantata 156)
III. Presto

These works, plus the two remaining double concertos 1060 and 1061, are considered
in this final chapter dealing with the concertos.

The absence of either an original model or a parallel version with which to
compare these concertos obviously limits the extent of conclusions. Despite such
limitations, however, significant deductions may still be realized by comparing these
works with concertos already reviewed. Three initial questions which invite
consideration are 1) whether these concertos are indeed transcriptions; 2) the original

solo instrument; and 3) the likely original key.

200

201

Given the lack of parallel versions, it is not unreasonable to question whether
these concertos did in fact exist previously in another form. The autograph itself gives
some clues. The appearance of the text of both solo concertos agrees with that of the
other concertos in P 234 in that the ripieno string parts are a clean copy, indicating
transferral from some other source, while the continua and cembalo parts were added
subsequently and are at times heavily reworked. As in the other concertos, this process
is evident from the bar lines, which typically are scribed for the ripieno parts first, then
extended downward to accomodate the continua and cembalo staves. Occasionally the
more florid cembalo discant spills over the bar line into the next measure. Thus while
the manuscript again appears to be a working copy, it is evident that certain parts were
not new but derived from other sources. Further, the existence of a parallel version for
the second movement of 1056 also suggests the existence of another version, even though
a parallel source for one movement does not necessarily guarantee the existence of
models for other movements. Given the appearance of the autograph, however, there
is no dissent among scholars that these concertos, like their companion works, are
transcribed from earler, now lost, models.

This judgment is also supported by the character of some of the solo writing
itself, which sometimes contains melodic patterns that betray idiomatic string figurations,

as can be seen in the first movement of 1056 (Ex. 5-1).

202
Ex. 5-1. 1056 I, 60-62, 92-95

f . A

  
  

60-62

 

92-95

 

V \V’ \—/

Similar violin figuration in mm. 47-54 of the same movement led Wilhelm Rust not only
to the conclusion that the original setting was for violin, but also was in the key of G
Minor, since these measures take advantage of the open G string. Fischer’s
reconstruction in the NBA of this work as a violin concerto in G Minor thus

acknowledges Rust’s hypothesis (Ex. 5-2).

203
Ex. 5-2. 1056 I, 47-54

---—-‘-_-'Al—VL'__1—V——V—V— .
~.-—- -~-——-- - u--"~u-- - I... "-1 _ _1 u- _--
-————-‘----—— a!

____—_-——_— l

--4 '----- .- .JH:_—4v‘.—--—‘-—_—a.v I 'n---- n -
=.--.w-—_—-----m-l-n----—-
, 7 7 a \ N‘ “; . ”_—

 

 

 

It should be noted further in this example that Bach retained verbatim the wide-spaced
violin figuration in the cembalo arrangement but divided the passage between the hands
to accomodate what would otherwise be an ungainly technical obstacle for the
harpsichordist’s right hand: rapid jumps of up to an octave plus a seventh. At the same

time, this modification also provides opportunity to add harmonic filler into the cembalo

 

 

204

bass, particularly desirable in this passage since the ripieno strings play only long
sustained tones. Although this movement betrays figuration idiomatic to the violin, it
is not so inherently violinistic as to require significant adaptation to the keyboard, as was
true in 1054. The accomodation of the wide skips in mm. 47-54, achieved without
rewriting the passage for cembalo, is but one illustration. When transposed back to G
Minor, the original key as suggested by Rust and others, the resultant range of the solo
violin is g-e“”’ which agrees fully with the typical range of Bach’s solo violin writing
in the concertos.

Determination of the original solo instrument in 1055 is more problematic.
Friedrich Spiro was among the first to suggest that 1055 was not originally for Cembalo,
but for a melody instrument.‘ He also postulated that the unusual arpeggios which begin
the first movement ritornellos were not original but later insertions,2 and, based on the
instrumental technique of the solo episodes, suggested that the original solo instrument
was the violin. Although other instruments have been proposed since, including the
cembalo or organ,3 Sir Donald Francis Tovey’s thesis that the original solo instrument
was not violin but oboe d’amore, as based on style and range considerations, has gained
broadest acceptance. Siegele’s further careful study of range and parallel passages in

1055 supports Tovey’s hypothesis, and this instrumentation is reflected in Fischer’s

 

‘Friedrich Spiro, 'Ein verlorenes Werk Johann Sebastian Bachs," Zeitschrift der
Internationalen Musikgesellschaft 7 (1905): 100 ff.

2Three of these arpeggiated passages--mm. 10, 24, and 42--appear as corrections in
P 234, confirming that they were worked out at the time the harpsichord concerto was
arranged.

3Aber, "Studien, " 12.

205

reconstruction in the NBA.4 That Bach was familiar with the oboe d’amore is evident
from its use in many cantatas, particularly in cantata choruses and arias which exploit
similar virtuoso capabilities as found in this concerto. The first known use of this
instrument in Bach’s works is from 1723 in Cantatas 75 and 76, but earlier incorporation
in lost works dating to Gothen may be possible.“ The range of the oboe d’amore found
in these works, a-b”, is also virtually the same as that found in 1055 if the interpolated
arpeggios in the cembalo discant are excluded. This range is easily accomodated by the
cembalo, and the fact that there appear to be no corrections in the autograph due to key
transposition leads to the judgement that the original key of the model was probably also

A Major.

Performance Questions: Figured Cembalo Parts

Perhaps the most intriguing feature of 1055 is the existence of a set of autograph
parts for this work, the only surviving set of original parts for any of Bach’s cembalo
concertos. These parts were available at the time Rust prepared the BG edition, but
were lost for nearly 35 years after the Second World War, coming to light again only

around 1980.“ Of greatest potential significance is the inclusion of a separate figured

 

- 4Siegele, Kompositionsweise, 129-130. A more complete discussion of fine points
relevant to the original instrumentation of 1055 is unnecessary here, but may be found
in NBA KB VII/7, 63-67.

“NBA KB VII/7, 65.

“Wolff, ”Leipzig Chamber Music,” 170. The set of parts, Mus. ms. Bach St 127,
though cited in the NBA KB, was still unavailable when series VII/7 of the NBA was
completed (1971). Rust, in his preface to the BG, acknowledged that the violin, viola,
and figured bass parts were from Bach’s hand, while the violone is "partly autograph"
and the cembalo certato ”van zweifelhafter Authentizitat." See BG 17, xviii and NBA

206
continua part, which suggests the presence of a second harpsichord to fulfil the

traditional role of basso continua (Fig. 5-1).

- _—-"".'=' .'Zu:='n_' -—--
v‘ ‘ :. - :=:" u.‘_.:

“—--.-=:E—l-nv—;===QI--r“= _.===_-:.-

s —--=---u --"— “a:
_._‘w _u— — r- .-
—. .l:— -—

     
   
 

 
    
 
 

-=:'=.-—'.' :"_'.':.'='.-E_::=: : :=_—‘Efi§;' __..-._
n._. . ;. ._ L_ 2. ..
::=:.__-.:=:.-.:. :: _ t a
o. .. a . 7‘ . -—“ .
’35-“? ti,- ‘~ -..-.~.'-;
~_ .

 

a:=‘:..::£fi§=vav. i—‘u-u I ”T... - n. 4

Fig. 5-1. 1055 I, II, Continua part from St 1277

 

KB VII/7, 67-68.

7Reproduced from Wolff, "Leipzig Chamber Music," 171.

207

St 127 also contains a separate violone part which doubles the continua bass, but
only in tutti passages. Thus continua and violone are distinguished from each other, at
least to the extent that the continua part is figured (implying a keyboard instrument to
supply harmony) and the violone plays only during tuttis. Wolff suggests that the
appearance of a second harpsichord to furnish the harmony plus the violone doubling the
bass should be taken as normative for all of these concertos.“ This is a question which
is not easily resolved, if it is answerable at all. As noted earlier in chapter 4, the
autograph score of 1057 also contains references to a violone part, but the intended
meaning remains clouded. Spitta affirmed the idea that a second keyboard instrument
was employed in these concertos; but having so acknowledged this inalienable principle
of baroque performance practice, he went on to acquiesce in the face of certain
conditions:

The part filled by this second harpsichord in the way of supporting the harmonies

is generally so slight that it could easily have been undertaken by the solo

instrument, and, indeed, it seems to me that in the last recensions of the D minor

concerto (1052) and the G minor concerto (1058) it was intended that this

alteration should be made.9
Spitta’s observation is not insignificant, for it highlights a judgment which no doubt
eventually brought about the discontinuance of continua practice: the recognition that the

continuo’s traditional functions of completing the harmony and reinforcing the bass were

already duplicated elsewhere in the musical fabric. Indeed, in those cembalo concertos

 

“The opposite view is held by Fischer: ”Die fiir das A-Dur-Konzert verbfirgte
Mitwirkung eines eigenen Continuo—Cembalos ist ja durchaus nicht ffir alle
Cembalokonzerte Bachs gesichert. " ("The participation of a separate continua-cembalo
as established for the A major concerto is by no means assured for all of Bach’s
keyboard concerti." My translation). NBA KB VII/7, 87 fn. 22.

“Spitta. Bach Ill, 140-141.

reg
it. .-
solo
(with
the vi.
the rip j

the dif

208

where the cembalo bass carries the original continua part and chordal filler is also added,
this duplication is especially apparent. It is notable that the two concertos singled out
by Spitta are the very ones in which only modest amounts of chordal filler are added and
where the cembalo bass is most independent from the continua part. It is rather in
concertos such as 1053, 1054, and perhaps 1057 that the cembalo is closer to a continua
function than in the works cited by Spitta. If indeed Bach intended a second harpsichord
for all of these concertos, then there would seem to be little purpose served in adding
additional chordal filler to the solo cembalo part. As already noted in 1053. the
harmony added into the cembalo part in certain passages of that work is an exact
realization of the figured bass which appears in the parallel cantata version. If this
harmony was already supplied by a second harpsichord, its duplication in the solo
cembalo would appear to have little merit.

Such considerations, however, do not fully resolve the issue in the face of the
textual support provided by St 127. It is particularly unfortunate that autograph parts for
the other concertos do not survive, since their content would be especially relevant to
this question. It remains possible that a second harpsichord was employed as the
requisite continua instrument for the ripieno in contrast to the solo instrument set against
it. In this scenario, the concertante cembalo occupies the same place and function as the
solo violin in a violin concerto. While the functional distinction between solo and tutti
(with its own continua) is retained, there is nonetheless a momentous difference: unlike
the violin in a violin concerto, the distinct timbre of the solo cembalo is mitigated when
the ripieno also contains a harpsichord playing the same harmony. Spitta also recognized

the difficulty with this arrangement: that with the presence of an accompanying

 

ml.

with

IEprO‘
(Iii/gr,
anothe
Wiser
bass Ii

209

harpsichord. "there could be no sort of distinct opposition between the tutti and the solo
instrument, either externally or internally."10 Such lack of timbrel and functional
distinction may open the way further to abandoning the continua, conceivable in a
concerto for harpsichord but not in a concerto for violin. Bach’s manner of handling
solo episodes in these harpsichord concertos implies that he was not unaware of this
problem of timbrel sameness, for not infrequently the continua part during solo episodes
is significantly thinned if not abandoned altogether, allowing the solo cembalo to sound
unencumbered. Indeed it is during the cembalo’s most brilliant solo episodes in 1055
that the continua is tacet. As illustration of this, one need only examine the first
movement of 1055. In each solo episode, the continua part either drops out altogether
(mm. 20-22, 25-32, 54-56, 69-72, and 77-78), or is very thin. It is also quite notable
that the continua part in St 127 includes the notations solo and piano at the beginning of
every solo episode,11 further suggesting that the continua player was to yield to the solo
cembalo in solo passages.

The concept of two keyboard instruments in a single work is perhaps not
altogether foreign to Bach. Laurence Dreyfus in his monumental study on Bach’s
continua practice examined the controversial subject of double continua accompaniment

with organ and harpsichord in some of Bach’s cantatas and passions. Citing

 

”Spitta, Bach, III 140.

11These directives are found at mm. 18, 19, 42, 51, 59, 67, 73, and 81, and alternate
with Tutti marked at the return of each ritornello. These performance signs are
reproduced in the BG edition, and clearly were taken from St 127 as the solo
designations do not appear in P 234 (some of the pin markings in P 234 appear to be in
another hand, and possibly added in later). The indication spicatto which appears at the
outset of the continua part in St 127 supports the hypothesis that the cellist doubling the
bass line played from the continua part.

210

documentary evidence of autograph parts which are clearly marked Cembalo, as well as
contemporary accounts which refer to the use of both harpsichord and organ in at least
some performances of Bach’s church music, Dreyfus concludes that dual accompaniment
was used at least intermittently throughout Bach’s Leipzig years.12 While care must be
exercised when drawing conclusions concerning the performance of secular instrumental
music from practices employed in sacred choral music, it nonetheless remains that the
notion of two keyboards, each furnishing the same harmonic material, may not be

foreign to musicians of the later baroque.

Derivation of the Cembalo Bass

The absence of parallel versions or extant models for 1055 and 1056 clouds
determination of the relationships between cembalo bass, continua bass, and the original
continua from the lost model. In some respects, these two concertos are simpler in
compositional approach and structure than some others. In both works, the cembalo bass
and continua parts are mostly doubled in tutti sections, but differences may be found in
the solo episodes where the cembalo bass often assumes greater rhythmic and harmonic
interest. When reconstructing 1055 as a concerto for oboe d’amore, the editors of the
NBA derived a continua part directly from the basses of 1055: whenever the cembalo
bass and the continua double each other, as in ritornellos, this same bass was retained,
but in solo episodes the cembalo bass was used except where the autograph betrays

corrections in the left hand part, in which case the reading before corrections was

 

l2Dreyfus, Bach’s Continua Group, 32-71.

the Pill

211

adopted.13 Thus the cembalo bass was judged to be the descendent of the original
continua part, i.e. , the carrier of the original bass. Accepting this assumption, however,
means that the cembalo bass in 1055 is virtually unaltered from the supposed original
reading, a rather unlikely event and a circumstance not found in any other concerto in
P 234. Such a hypothesis would mean that Bach made no changes to the continua part
when adapting it for the left hand of the cembalo. Based on the pattern of the other
concertos, it would seem more probable that Bach started from a simpler bass line in the
continua and at least occasionally embellished it for the keyboard bass. In the following
passage, for example, it is likely that the original continua was more skeletal, as in the
concerto continua, or at the least less florid than the transcribed cembalo bass, as

suggested below (Ex. 5-3).

 

13There are very few such corrections discernible. mm. 71, 75-78, and 87-88 being
the principal examples.

Ex. 5-3. 1055 1, mm. 62—65

 

 

 

. l

. — r— l _.—— _ r— —_._
...- __ ___—-_.—_ .— ‘_—. _——_————-. ._-'—-'— h—
.u—n—_—l-_———-_——‘ - ———- —_ i-______

I _ ___ ———-—

Im o .-
..‘v--—"-1—_n-‘nn—n-4r‘—r——_ :-.-._—-r—-‘-

g..- n...=====.. c::----- “4..., ._ ._ _._.... -__:-

. .- ——-— —— _—

r -_-- In“: -- —— -

 

 

 

Similar observations are applicable to 1056. While the two bass parts again mostly
double each other, there are passages in which greater differentiation prevails, and the
NBA reconstruction reflects this by sometimes favoring the 1056 continua part over the

cembalo bass when reconstructing the original continua, as in the following examples

(Ex. 5—4).

 

I” the
adapt

213
Ex. 5—4. 1056 I, 39—42

-T—\

_.— \—
, i - —- .m—r-
\ _-=== S..—* 1.--“:

 

1056 1, 96-100. reconstruction

    

. 0 .—
._,,__-,_— —...—— —__————.
u..- —— '

_-

 

In the last example, the long sustained C, unfeasible on the harpsichord, necessitates
adaptation, as does a similar passage at 47-55. In the latter instance, not only the left

hand requires accommodation, but also the wide leaps of the discant (which betray the

214

string provenance of the work) present a technical hurdle which is resolved by dividing

the passage between the two hands (Ex. 5-5).

Ex. 5—5. 1056 I, 47-55

 

 

 

f8

215

Bach’s adaptation of this passage becomes even clearer when compared with the parallel
segment at mm. 92-94 (Ex. 6-6), where the contour of the discant is the same but the '

narrower leaps are playable under one hand.

Ex. 5-6. 1056 I. 92-94

 

Both of these concertos make almost no use of chordal filler; in 1055 there is none at
all, unless the brief appearance of double 3rds in mm. 67-69 is considered, while in 1056
there are but four instances, three of which are in the third movement.“ In essence,
then, both of these concertos are probably very close to their original models, with the
left hand modestly more independent of the original continua in 1056 than in 1055.
When compared with other concertos such as 1052, 1058, or 1062, however, both these
works appear conservative and less adventuresome, particularly in the handling of the
cembalo. This observation is reflected in the appearance of the autograph, for there are
comparatively few changes or corrections in the cembalo parts. Of the changes which
do appear, most are in the discant, not the bass. The A major concerto (1055) is
particularly clean, again suggesting that it is very close to its original model. As Spitta

remarked, these concertos ”elucidate the structure of the older form of concerto.”

 

1“1 mm. 49-54; 111 mm. 23-24, 99-104, and 203-209.

”Spitta, Bach III, 142.

216

Double Cembalo Concertos without Models

Besides the seven solo concertos contained in P 234, two additional double
concertos are worthy of brief review. The C Minor Concerto for two harpsichords,
1062, has already been considered, and is the only double concerto with an extant
model. It is also the only double concerto with a surviving autograph. The two remaining
double concertos, 1060 and 1061, have neither extant model nor parallel versions, thus
limiting the extent of conclusions that may be drawn. Nonetheless, there are some
corroborating observations to be realized. For purposes of the present study, the
concertos for three and four harpsichords are not Considered, since many questions
remain concerning the original models and authenticity of the triple concertos, and 1065
for four harpsichords is based on a work of Vivaldi}; Autographs also are lacking in
each case.

The Concerto in C minor, 1060, is less well-known than its companion work in
the same key (1062), since the latter has gained considerable familiarity as the popular
D minor concerto for two violins. The absence of parallel or model versions, as well
as the loss of original autograph materials, allows only hypothetical conclusions to be
drawn about the origins and transcription techniques used in this concerto. Rust noted
an analogous similarity between this work and the concerto for 2 violins in D minor
(1043), and, from the solo parts in the second movement and the idiomatic string writing
in the finale, concluded that the original form of the work was a concerto for two

violins. Waldemar Voigt and others following, however, while recognizing the

 

1“A quite detailed discussion of source and authenticity issues related to the concertos
for three and four harpsichords may be found in NBA KB VII/6.

17‘
and Sic
381m)”

217

violinistic figuration in the first cembalo also noted that the discant of the second
cembalo differed markedly, often favoring a cantilena style. This difference becomes
especially apparent in parallel solo passages. The inevitable conclusion drawn by Voigt
was that the original form of 1060 was for two different melody instruments, with violin
and oboe the most likely.17

Wilfried Fischer’s reconstruction of this work for the NBA as a concerto for oboe
and violin is derived very closely from Bach’s arrangement. The ripieno strings, as well
as the continuo line, are replicated as written; the cembalo I discant becomes the solo
violin part, while the cembalo II discant is fashioned directly into the oboe part. It is
altogether conceivable that this disposition of parts reflects quite accurately the original
reading, for the other concertos verify that Bach’s common practice was to make few if
any changes in the ripieno parts, and the two solo parts seem very well suited to their
respective instruments. The idiomatic distinction between the solo instruments is

especially clear in the third movement, where the violin figuration is obvious (Ex. 5-7).

 

1"Voigt’s observations subsequently were enlarged and corroborated by Waldersee
and Siegele, and also became the basis for several reconstructions of 1060 by Max
Seiffert, Max Schneider, and Wilfried Fischer. NBA KB VII/5, 35, and VII/7, 100.

218

Ex. 5-7. 1060111, mm. 125-130

 

 

 

 

 

____.

‘— —
_— _—--——_-
{-2,—_._.- ____—__

_ ——__—. ,
lf..-.—.—— _-_.:-
l--_—___.. —_——..

, _ l
E __ .‘-_-
I: .:.~,‘:———._—_:.——§—_——='_—.__——-—_='

. . .—___ _._.—___.._——
l: .-_-.._——. _______—___.—,=
——P_ =——__
=-, — —__
\——’
P 5"

I

 

 

Given the lack of autograph source materials, it is impossible to know what
changes or corrections Bach made when transcribing 1060. As Fischer comments in his
notes to the NBA, Bach seems to have stayed unusually close to his model, not even

adapting idiomatic string figures to keyboard as was generally his custom.18 If

 

l"Fischer, NBA KB VII/7, 101.

219

Fischer’s reconstruction is reasonably true, then Bach’s corrections and alterations may
well have been almost exclusively concerned with creating left hand parts for the
harpsichords. It is on this point that the greatest affinity between 1060 and 1062 may
be recognized: both works present in equal measure the problem of deriving two viable
left hand parts from a single original continuo bass, and in both works a similar end is
realized. While the cembalo basses generally double the continuo bass in ritornellos,
considerable freedom and exploration of keyboard figuration characterizes the solo
episodes, so much so that the left hand parts often show greater independence from the
continuo than is true in many of the solo concertos. Further, the cembalo II bass is often
more brilliant than the bass of cembalo I, the latter keeping closer to the continuo bass.
The following excerpt well illustrates this procedure, especially since the passage is from

a ritornello which more commonly is set with all three bass parts doubled (Ex. 5-8).

220

Ex. 5-8. 1060 1, mm. 22-29

 

 

yu—
an.
'-

 

 

‘—_' _
vu—__———— —— __-.
'- —

_.‘--

r~\——— n-—-——

“tum—- ———a—_

r-—-————_— — - —

_fi———.———
-

  

 

'~ '—~_—‘-—————‘—
.1 In.

_._.—.— 1-
——__ -77 7‘--
fi————

 

  
 
  

a-—— .--.-_-_ — .‘-
_——--———-— ---——
— _-_- —
——-—

  
   

 

 

221

One might further speculate that the discant of the second cembalo, being drawn from
the oboe part and thus favoring cantilena over brilliance, offers greater opportunity for
the bass to be more brilliant as compensation for the less brilliant discant. Lastly, it
might be observed that there is a considerable amount of chordal filler added into the
first movement, but none in the remaining movements. This circumstance appears to
bear no special consequence.

Rather little space need be given the third double concerto, 1061 in C Major.
This work is so unlike any other that its very inclusion among these harpsichord
concertos raises numerous questions. From the outset of the opening movement, it
becomes immediately apparent that the relationships between the two solo harpsichords
and the ripieno strings are altogether different than in any of the other concertos. As
noted in the latter, the opening tutti always finds the solo parts doubling the ripieno

through the exposition of the first ritornello (Ex. 5-9).

Ex. 5-9. 1061 1, mm. 1-10

 

222

 

Here the orchestra is completely ancillary in function, without the usual contrast between
tutti and soli, so that the usual concerto dialogue is absent. The strings are never truly

obbligato, but merely double the keyboard parts lightly, or play simple chordal

223

accompaniment for color. The most obvious manifestation of this changed role of the
ripieno is the very lengthy solo episodes during which the strings and continuo are tacet,
leaving the two cembalos to a rich musical dialogue of their own. Indeed, the second
movement is played entirely without the ripieno, and in the third movement, a thorough-
going fugue, the first violins do not enter until m. 33 nor the violas and continuo until
m. 45, which is exceedingly late (consider the absence of the continuo for 44 bars). Not
only are the harpsichords totally dominant, but equally insignificant is the role of the
continuo, which is silent through 60 of the 166 bars comprising the first movement.
As early as 1802 Forkel suggested that 1061 could be played successfully without
the accompaniment of strings and still sound most admirable.19 The non-obbligato
strings, unique design, and extant sources all suggest that this work may originally have
been conceived as a duet for two harpsichords, later supplemented with a thin string
accompaniment. Although no autograph sources survive, it is curious that the source
closest to Bach is a set of original parts for the two harpsichords--without parts for
strings--in the hand of Anna Magdalena (Mus. ms. Bach St 139), with title, movement
headings, Adagio marking at I 165, and ten bars of the first cembalo part (mm. 100-109)
in Bach’s own hand. It is notable that each cembalo part in this manuscript bears the
title in Bach’s hand Concerto. l a due Cembali. : di 1 J. S. Bach. Compared with the
titles of the other concertos, where the complete instrumentarium is listed, the strings are

conspicuous by their absence. Several later manuscript sources for either or both

 

19Johann Forkel, Johann Sebastian Bach: his life, art, and work (New York: Da
Capo Press, 1970), 131.

224
harpsichord parts also survive, and have been grouped together by the editors of the

NBA as materials connected with a ”version” for keyboards alone.20

For what purpose Bach may have composed such a unique work, if indeed for
harpsichords alone, is unknown. Spitta had dated the work to 1727-1730,21 but the best
source, St 139 cited above, bears the "large MA" watermark which dates to 1732—35.
Since it is likely that the double and triple concertos were written to be played by Bach
and his sons in the Collegium’s Ordinaire concerts, and Emanuel and Friedemann were
in Leipzig only until 1733 or 34, it is probable that 1061 was composed sometime
between 1729 and 1733, during Bach’s first period of association with the Collegium.
More perplexing is the nature of the work itself. Is it really a concerto, a counterpart
to the Italian Concerto for solo harpsichord without orchestra, or rather a duet for two
equal instruments? Given the minimal appearance of the continuo part, any question of
a third harpsichord to play continuo bass seems immaterial. By any standard, 1061
seems to belong more to the realm of solo chamber music than orchestral chamber
music, particularly given the intricate contrapuntal interdependence of the two
harpsichords. The emphasis on counterpoint is unmistakable: the middle movement
explores the possibilities of strict four-part polyphony, while the finale is a remarkable
and highly crafted fugue not far removed from Bach’s other keyboard fugues. Spitta’s
comment about this fugue seems particularly apropos:

The fugue belongs to the sonata form, or to that of the concerto in the sonata

style; it has nothing in common with the strict concerto form, since that originates
not in polyphony, but in homophony, and its working-out is not thematic but

 

20Karl Heller and Hans-Joachim Schulze, NBA KB VII/5, 75-85.

21Spitta, Bach, III 144.

225

episodic. (In this movement Bach) succeeded in a most masterly way in suiting
the form to the character of the movement, by the style of invention and
treatment, especially by means of longer episodes, or even interludes, quite in the
free style; and he was led to introduce them by the style of the harpsichord and
the organ, which always influenced his imagination.”
The notion of a "concerto in the sonata style” is intriguing, for it suggests the blending
of chamber and orchestral styles, much as the Sonate auf Concertenart to be explored
later. And further, the renewed application of strict fugal principles within a larger
multi-movement work is something Bach returns to elsewhere in his late works,

particularly in the ricercar of the Musical Ofi’ering and the mirror fugues in the Art of

Fugue.

 

22Ibid. , III 145-146.

226

Summary of Observations from the Concertos

The foregoing chapters have examined the seven concertos for solo harpsichord,
1052-1058, the concertos for two harpsichords, 1060-1062, and all the related extant
models or parallel versions which survive. The following summarizes the principal
conclusions.

1. As best as can be ascertained, these concertos all stem from the years of
Bach’s association with the Leipzig Collegium (1729-37 and 1739-41). There is broad
agreement among scholars that these works were written for performances of the
Collegium, with the multi-keyboard concerti appearing first, most likely while Emanuel
and Friedemann were still part of the Bach circle in Leipzig. The Collegium provided
the perfect workshop for proving this new genre, and it is possible that at least some of
the changes appearing in the autograph stem from hands-on experiences of rehearsing
and performing these new works.

2. The appearance of the autographs gives testimony to Bach’s compositional
process. The ripieno string parts were written in first, and usually are transmitted as a
clean copy or with relatively few and minor changes; the latter, if required, are usually
necessitated by alterations in other parts. The continuo bass and harpsichord parts were
copied subsequently, with the two bass parts often worked out simultaneously. The most
significant changes appear in the cembalo, both discant and bass. Several types of
corrections may be distinguished, including accomodation of string idioms to keyboard,
alterations for range, arpeggiation, embellishment and ornamentation to achieve greater

brilliance, added chordal filler, and exchange of parts.

227

3. Inasmuch as these harpsichord concertos are all arrangements of earlier
concertos, Bach already was well acquainted with the concerto genre. And since several
of these works had already been arranged previously from (lost) models into another
form, e.g. , church cantata movements, the very process of transcription was not new,
but already practiced in Bach’s hands. Thus these concerto transcriptions were not first
attempts or experiments in transcription. Further, since the multi-keyboard concertos
appeared first, Bach had gained additional experience dealing with idiom and timbrel
issues before setting out to write concertos for solo harpsichord. The latter comprise
Bach’s final works in the concerto form, works which may be seen to culminate Bach’s
life-long cultivation of the concerto.

4. In almost all cases, the cembalo concerto appears to have been transposed
down a step, most likely to accomodate the range of Bach’s harpsichord, which
apparently had a maximum range of GG to d’”. However, since at the time the ambitus
of harpsichords varied from instrument to instrument, it is also possible that some
concertos presumed a different instrument of more limited range.

5. The alterations and corrections found in the keyboard parts are consonant with
the kinds of revisions Bach made in all his keyboard music, i.e., the reworkings
encountered here are not unique to the concerto repertory alone. George Stauffer has

identified several distinct types of revisions found in Bach’s solo keyboard repertory:23

 

23’George Stauffer, ”Bach as Reviser of his own Keyboard Works, " Early Music Vol
13 N0 2 (May 1985): 189-197.

228
a) Changes of detail, usually involving modification of figuration or arpeggiated

patterns. In the concertos, adjustment of string figuration to a keyboard idiom is
a common example (e.g. 1054 I, 57-69).

b) Expansion of selected bars, thus extending a passage by inserting additional
measures. While less frequent than (a) above, examples do occur, as in 1053 I, 13.
c) The complete reworking and expansion of an existing piece. In a sense, all of
the harpsichord concertos incorporate this process, particularly in realizing a new
cembalo bass. The second movement of 1053 also illustrates this practice with a
free extension of material from m. 13 while omitting measures elsewhere.

6. Of greatest interest to the present study is the formation of a viable left hand
part from the continuo bass of the model, an important process requisite to the
emergence of the cembalo as a solo rather than continuo instrument. Siegele finds
Bach’s procedure divisible into two basic types?"

(a) The first is that in which the original bass is taken over directly into the cembalo
bass, the new continuo part then either doubling the left hand part, particularly in
ritornellos, or following the outline of the cembalo bass in a more simplified form.
In unaccompanied solos, the continuo part may drop out entirely, leaving the left
hand to carry the original bass alone. Generally, the continuo takes over the
original bass only partly, while the left hand retains it complete. Siegele cites 1054

and 1057 as examples of this paradigm.

 

24Siegele, Kompositionsweise, 125.

229
(b) Siegele’s second type finds the original bass transferred principally to the

continuo, the new cembalo bass then either doubling this continuo line or playing
a free part which embellishes the contour of the continuo or plays more independent
figures. The latter is most common in solo episodes, but also occurs sometimes in
ritornellos. Thus the continuo is the principal carrier of the bass. Siegele suggests
1058 and 1062 as examples of this type.

While Siegele’s categorizations are interesting and somewhat helpful, they also
represent an oversimplification, for there is considerable variety in the relationships
between the three bass parts in these concertos; in some cases, the relationships are not
even the same in all three movements of a given work. Furthermore, Siegele raises the
question of which type represents a stronger form of transcription, suggesting that 1052a,
presumably produced from the less practiced hand of Emanuel, illustrates the first type
while Sebastian’s more strongly worked out 1052 represents the second.25 Implicit in
this dichotomy is the notion that the second type, in which the cembalo bass is more
independent, exemplifies a more developed type of transcription, and hence the works
so constructed belong to a more advanced stage in the development of the keyboard
concerto. This reflects more a historical than contemporary viewpoint, however. There
is nothing to suggest that Bach viewed these works with any prejudice, nor that the so-
called type I works are any less viable musically. The difficulty of imposing categories
becomes especially evident in the case of a work like 1057, which Siegele assigns to type
I, ostensibly because the original continuo is present more in the cembalo bass than the

continuo. While the left hand part does indeed carry the original bass often (usually

 

2"’Ibid. , 125-126.

230

during ritornellos), there are also frequent instances where the cembalo bass is so heavily
ornamented that the bass really is clearer in the continuo, or, even more significant, a
few occasions where the left hand takes up thematic material from the discant, a rare
occurence in these works.26 The inclusion of principal thematic material in the cembalo
bass advances the importance of this part, regardless of where the original bass may lie.
The cembalo bass is tied so consistently and closely to the original continuo that even
a brief excursion into musical material not from the borrowed bass becomes significant,
widening the gap between solo and continuo. It may be more accurate to suggest that
the independence of the cembalo bass is due as much to the number of parts as to the
role of the continuo: 1057, 1060, and 1062 all show a tendency toward greater freedom
of the left hand part, and all have extra instrumental parts in the texture (two recorders
in 1057, a second harpsichord in 1060 and 1062). On the other hand, 1058 and
especially 1052 illustrate concerti in which the left hand exercises considerable freedom
from the continuo despite the more standard instrumentation.

7. The varying amount of chordal filler present in these concertos does not
appear to follow any particular pattern, at least in relation to the independence of the
cembalo bass. Very little if any chordal filler appears in 1055 and 1056, a fairly
considerable amount is incorporated into 1057, and the remaining works fall somewhere
in between. Interestingly, chordal filler is not found in middle movements except in

1057 and 1058, both of which contain significant amounts. While added chords may

 

2"’Examples include I mm 53-56, 63-64, 125-137, 241~243, 263-271, 311-325. The
finale, being fugal, also contains examples. Mm. 263-271 are particularly notable, since
the original continuo bass drops out entirely for eight bars while the cembalo bass takes
over the theme of the violino principale from the Fourth Brandenburg model.

231

occur in either the cembalo bass or discant, they are more common and frequent in the
discant. This observation is of some significance, since the discant is where the continuo
chords were customarily placed in Bach’s time when realizing a thoroughbass.27 Thus
added chordal harmony, particularly in the discant, may be seen as vestiges of the
cembalo’s older continuo function, and the rhythmic placement on principal beats plus
the close position placement of chord tones further supports this idea. Siegele observed
that the cembalo is most emancipated from its traditional role of thorough-bass
instrument in those works where the continuo carries the bass and the continuo function
is maintained only in tutti passages.28 While chordal filler is perhaps more common
in tuttis, it is by no means limited to tutti passages but may also be found during solo
episodes as well.

8. Unanswered questions necessarily remain, in part because of the lack of
original source materials, both for parts and lost models. Especially vexing are questions
relating to the use of a second harpsichord as continuo instrument and the inclusion of
the violone in the instrumentarium. Support can be garnered on both sides, but in the
absence of autograph parts it is likely the issues will remain unresolved. It is curious
that the one concerto for which parts do exist (1055) and which includes a figured bass
part for continuo is also the only solo harpsichord concerto which has no chordal filler.
Could it be that there was no need of chordal filler given the addition of a second

harpsichord to furnish the continuo harmony in this particular work? Or, put the other

 

27See The Bach Reader, 388-390. The example of a thoroughbass realization by
Kimberger from his Grundsatzes des Generalbasses of 1781, reproduced on p. 388,

well illustrates this custom.

28Siegele, 127.

232

way, that with the inclusion of chordal filler in many of the other concertos, there was
no need for a second harpsichord to function as continuo.

9. The larger issue of the relationship of these concertos to the history of the
keyboard concerto is not really the focus of the present study, but it is a question which
inevitably is raised by scholars dealing with these works. While a direct line of descent
may be difficult to document, several factors do suggest that these concertos were highly
significant in the blossoming of a new genre. The further development of the keyboard
concerto leading to Mozart was through CPE and JC Bach, both of whom had to have
been influenced considerably by their father. As already noted, CPE clearly had a close
connection with 1052a, since five parts for this work survive in his hand (St 350).
Perhaps this arrangement was a pedagogical assignment from Sebastian which served to
introduce the younger Bach to the possibilities inherent in the new genre.

At least some conclusions can be drawn, however. Certainly Bach’s keyboard
concertos mark an inauguration of greater import than the fifth Brandenburg, which
Besseler, Geiringer, and others hailed as the first keyboard concerto. The fact that three
instruments share solo honors in the fifth Brandenburg places this work in another
category altogether. Bach’s turn to the cembalo for the solo instrument, rather than the
more traditional string or wind melody instrument, marked a turn in a new direction.
The three-movement form common to baroque concertos was retained, as well as the
tutti-solo contrast and recurrent ritornello form. This is not Mozart’s concerto, but many
of the elements are there, even including the continuous participation of the keyboard
instrument. It may be recalled that the scores for many of the Mozart concertos call for

the inclusion of the piano as a continuo instrument during the tuttis, a practice which can

233

only be traced back to the baroque and to Bach’s concerto practice. To what extent this
later practice in Mozart may suggest a similar and accepted practice in Bach’s time is
unanswerable. However that may be, it remains reasonable to presume that the
connection with the Collegium awakened Bach’s renewed interest in the instrumental
concerto, and particularly in the harpsichord as a solo instrument, with the result that
these wonderful examples of Bach’s invention were brought to life. This creative focus
on writing for the cembalo was not limited entirely to the concerto, however, but also
finds precedents in the sonatas with obbligato harpsichord. The following chapters turn

to these significant compositions.

PART II: THE OBBLIGATO SONATAS

 

VI.

The Sonatas with Obbligato Cembalo

The harpsichord concertos examined in the preceding chapters are not the first
ensemble works for which Bach composed a solo cembalo part. The basso continuo had
been the common denominator in all styles of music in the first half of the eighteenth
century, but as the century progressed strict adherence to the conventions of continuo
practice began to weaken, and it was in chamber music especially that composers first
pioneered efforts to do away with the continuo. Bach’s contributions in this process are
represented by the following chamber sonatas with obbligato cembalo:

Six Sonatas for Violin and Cembalo, BWV 1014-1019
#1 in B Minor
#2 in A Major
#3 in E Major
#4 in C Minor
#5 in F Minor
#6 in G Major

Three Sonatas for Viola da gamba and Cembalo, BWV 1027-29
#1 in G Major
#2 in D Major
#3 in G Minor

Two Sonatas for Flute and Cembalo, BWV 1030, 1032
#1 in B Minor
#3 in A Major1

 

1The second flute sonata in E" Major, BWV 1031, and a sonata in C major, BWV
1033, both with obbligato cembalo, generally have been regarded as spurious works and
are to be included in the NBA in a separate volume so labeled. Robert Marshall,

234

235

Of the eleven sonatas listed above, one of them, the G Major gamba sonata, also exists
in a different version as a trio sonata, and it appears that the gamba sonata was derived
directly by transcription from the trio. In a process paralleling that found in the
concertos, the continuo bass of the trio is transferred directly to the cembalo bass in the
gamba sonata. Eppstein and others hold that several more of these sonatas, perhaps even

all of them, may have been derived similarly from now lost trio sonata models.
Chronology, Scope, and Sources of the Chamber Music

Before examining any of these sonatas, it is necessary to first consider the
chronological relationship of these works to the keyboard concertos. Despite extensive
efforts by leading scholars to establish a chronology for Bach’s instrumental oeuvre, this
most difficult issue remains largely unresolved. The considerable advances made in
Bach studies following the remarkable work of Dfirr and Dadelsen on the chronology of
the vocal works has not seen a parallel development for Bach’s instrumental output.2
Although more recent studies have been successful in determining manuscript scribes and
cataloging the watermarks of available materials, many questions remain. To a
considerable degree, this situation is due to the paucity of autographs which have

survived for the instrumental repertory, a condition which is in marked contrast to that

 

however, has argued that both works may indeed be by Bach, 1033 originally an
unaccompanied sonata to which the bass was later added by Emanuel, and 1031 as a late-
Leipzig work with galant elements. See Robert Marshall, ”J.S. Bach’s Compositions for
Solo Flute: A Reconsideration of their Authenticity and Chronology,” Journal of the
American Musicological Society xxxii no. 3 (1979): 464-473.

2Alfred Dfirr, ”Zur Chronologie der Leipziger Vokalwerke J. S. Bachs," B] 44
(1957): 5-162; and Georg von Dadelsen, Beitra'ge zur Chronologie der Werke Johann
Sebastian Bachs, Tilbinger Bach-Studien, Heft 4/5 (Trossingen: Hohner, 1958).

236

attending the vocal works. In most cases only scribal copies have survived. Among the
chamber sonatas listed above, only the G major Gamba Sonata (1027) and the two flute
sonatas (1030 and 1032) are extant in autographs, and most writers have considered these
works to be later versions of now lost models. With this absence of source materials,
scholars have been forced to turn to style analysis as the primary means of investigation,
often with inconclusive, speculative results.

Additionally, there is also a broader question concerning the size and scope of
Bach’s chamber music, an issue raised in the first chapter of this study which must be
reconsidered here. Even including chamber music with continuo, the number of
ensemble works is relatively small, prompting speculation that the extant works may not
fully represent the total sum of Bach’s output. As noted in chapter one, the earliest
recorded works list included as part of Bach’s obituary inventoried the six sonatas for
unaccompanied violin, the parallel six sonatas for unaccompanied cello, the concertos
for one to four harpsichords, and "a mass of other instrumental pieces of all kinds and
for all kinds of instruments."3 This broadly encompassing statement has been
interpreted as indicating that a larger corpus of such pieces once did exist but
subsequently has been lost. While some losses may have occurred, the catalog contained
in the obituary is not definitive nor exhaustive enough to make a firm judgement.
Among the list of published works, Cantata 71 and the Schemelli Gesangbuch are notably
lacking, and the number of pieces inventoried under "unpublished works" may be
overstated. The latter includes five full cycles of church cantatas and five passions.

That the third and fourth cantata cycles are incomplete and the fifth Jahrgang non-

 

3The Bach Reader, 221.

237

existent while the first two cycles are entire prompts speculation that five full cycles may
never have really existed. Of further note is Forkel’s annotated works list in his 1802
biography, which includes the citation "many single Sonatas for the harpsichord with
accompaniment of violin, flute, viola da gamba, etc., all admirably composed so that
even in our days most of them would be heard by connoisseurs with pleasure."
Forkel’s biography was written fifty years after Emanuel’s Nekrolog and clearly is
indebted to it at many points. Nonetheless, it is perhaps significant that Forkel’s list
includes flute, violin, and gamba, the very same instruments for which works survive
today. Forkel’s "etc." is perhaps a nod to Emanuel’s imprecise ”mass” citation, leaving
the door open to additional (unknown) works. While it is possible that already by
Forkel’s time only the sonatas for flute, violin, and gamba had survived, others having
already been lost, it is also possible that there never were additional chamber works for
other instruments.

Given Bach’s considerable use of the oboe in choral music, some scholars have
found the absence of sonatas with oboe particularly astonishing.‘ While it is quite
possible that Bach may have written chamber works for oboe(s), there is nothing that
necessitates such a presumption. In the light of this question, Alfred Wierich’s study of

the sonata for obbligato clavier with violin is most interesting.6 While Wierich’s survey

 

4Forkel, reprinted in The Bach Reader, 344.

”The oboe and recorder hold such an important place in the scoring of Bach’s vocal
works that their absence from his sonatas is simply inconceivable.” Wolff, "Chamber
Music,” 166.

6Alfred Wierichs, Die Sonate fiir obligates Tasteninstrument und Violine bis zum
Beginn der Hochklassik in Deutschland (Kassel: Barenreiter, 1981).

238

of sonata literature is principally directed toward violin sonatas, works for other
instruments are also included. What is notable is that among the repertories considered
from over forty German composers, the violin and flute are commonly represented and
viola da gamba also is encountered frequently, but oboe is noticeably rare. In other
words, the instruments favored for sonatas with obbligato cembalo are the same ones
found in the extant Bach oeuvre. Newman’s study of the sonata in the baroque yields
a similar picture: that the violin is the preferred instrument, and that oboes, even when
mentioned, are more often than not either substitutes for violin or appear in a mixed
instrumentation with continuo (not obbligato cembalo).7 It is beyond the scope of the
present study to pursue this question further, but initially it would appear that the oboe
as a solo instrument in chamber genres is not necessarily a common phenomenon in the
period. Thus it may be unwarranted to ponder an original chamber repertory of larger
size on the basis of absent oboe sonatas.

As the table of instrumental works on page 35 demonstrates, the majority of
Bach’s known chamber music is traceable only to sources dating from Leipzig. On the
surface, this observation appears to be at odds with the dichotomous view of Bach the
Cothen Kapellmeister versus Bach the Leipzig Cantor commonly presented in many
biographies. The assumption of such presentations has been that most of the chamber

music could only have emanated from C6then or even Weimar, since Bach’s duties at

 

7William Newman, The Sonata in the Baroque Era (Chapel Hill: The University of
North Carolina Press, 1985), 255-300. A further and fairly extensive catalog of oboe
literature draws the same observations: that oboe sonatas with continuo are relatively
few, oboe sonatas with obbligato cembalo don’t exist, and that the majority of baroque
works with oboe are either concertos or ensemble pieces of mixed instrumentation.
Bruce Haynes, Music for Oboe, 1650-1800. A Bibliography (Berkeley: Fallen Leaf
Press, 1992).

239

St. Thomas’ in Leipzig kept him preoccupied composing sacred music. Hence any
chamber music which surfaced in Leipzig could only be assumed to represent
arrangements or copies of works which originated earlier. While this hypothesis may
be true in some instances, it cannot be assumed in all. Certainly Bach’s duties with the
Collegium provided considerable opportunities for the performance of chamber music
beyond the keyboard concertos, and as noted earlier there is considerable evidence to
suggest that Bach continued to regard himself as a Capellmeister in Leipzig and that he
wrote and performed instrumental music even before assuming duties with the Collegium
in 1729. Ernst Ludwig Gerber, writing of his father Heinrich’s associations with Bach,
reported that already in the fall of 1724 his father had ”heard much excellent church
music and many a concert under Bach’s direction. ”8 While the venue for such concerts
is unknown, it remains quite plausible that chamber music may well have accounted for
a goodly share of the content, perhaps even presented by the Collegium prior to Bach’s
directorship.

Hans Eppstein’s extended monograph on the sonatas with obbligato cembalo
accepts and continues the commonly held view that the chamber sonatas or at least their
original versions stem from C(‘ithen.9 Eppstein acknowledges the dating of the Leipzig
sources, but in virtually every case argues for an earlier pre-Leipzig version--sometimes
a trio sonata, sometimes a concertouupon which the sonata is based. The exemplar for

this process is the G major gamba sonata, 1027, transcribed from the parallel G major

 

8The Bach Reader, 264.

9Hans Eppstein, ”Studien fiber J. S. Bachs Sonaten filr ein Melodieinstrument und
obligates Cembalo, " Acta Universitatis Upsaliensis: Studia musicologica Upsaliensia, 2
(Uppsala: Almquist & Wiksells, 1966).

240
Trio Sonata, 1039. Among all the obbligato cembalo sonatas, 1027 is the only work that

survives in a parallel trio sonata version. Eppstein’s conclusions about these works rest
on stylistic considerations and especially on close examination of the range of the parts
as a means to determine the original gestalt of a work. His premise is that when parallel
passages betray alteration to accomodate the range of the instrument, the work must have
existed originally in some other form, key, or scoring.lo To illustrate from these two
G Major sonatas, Eppstein has noted different readings involving octave transfers in

parallel passages, as in the first movement at mm. 11-12 and 23-24 (Ex. 6-1).

Ex. 6—1. 10391, 11-12, 23-24

11-12

 

23-24

 

The reading at m. 24 would appear to be the intended one, for in m. 12 the shift up a

seventh breaks the continuity of the line in order to avoid the low c#’ (the lowest

 

loSiegele applied somewhat the same analytical method to his examination of the
concertos.

241

available pitch on the transverse flute at the time was (1’).11 Curiously, at the same
. measures in the gamba sonata the octave shift is still preserved, even though range is not
an issue for the gamba, thus suggesting that the trio indeed was written first and the
gamba sonata was transcribed from it. Such is the traditional interpretation of the
relationship of these two works: the trio sonata preceded and served as model for the
gamba sonata.12 That transference from a trio to solo sonata was not unknown in the
18th Century (though principally by younger composers than Bach), rather than in the
reverse direction, also supports this view.

A third version of the first, second, and fourth movements of this work also
exists, apparently for pedal clavier or organ.13 There are some differences between this
reading and 1027/1039, including articulation marks that suggest wind or string
instruments and variants in the bass line which most likely are due to considerations of
range. The trio version (1039) in all likelihood is the earlier source, the keyboard
version being an arrangement from it. It is particularly noteworthy that while the
keyboard version (P 804) generally agrees with the trio, there are octave displacements
which follow the gamba version rather than the trio. This is particularly obvious in the

fugal fourth movement, mm. 93-97, where a characteristic ascending octave leap clearly

 

1‘The range of the transverse flute is verified by Mattheson, Walther, Quantz, and
Hotteterre; cf. H. P. Schmitz, Quenflote und Querflb'tenspiel (Kassel: Barenreiter, 1952),
26.

12This view has been accepted by Rust, Spitta, Schweitzer, Siegele, and most other
scholars.

13NBA KB VI/3, 50. The pedal keyboard version is contained in manuscripts Mus.
ms. Bach P 804, Mus. ms. Bach P 288, and Musikbibliothek der Stadt Leipzig Ms. 7.

242

belonging to the theme appears in the gamba and organ versions, but is omitted in the
trio version (Ex. 6-2).
Ex. 6-2. 1039/1027 IV, 93-97

1039

 

l 027

 

Such discrepancies led Eppstein and Siegele to the conclusion that neither the trio nor
the gamba sonata represent the original version, but that both hark back to an earlier,
now lost, original.“ Eppstein further suggests that the original version of the work
may have been for two violins and continuo, that 1039 was worked from this early
version by Bach himself, that 1027 is based on 1039, and that the keyboard version may

stand in some relationship to the early version as well."

 

l“Eppstein, "Studien,” 129; Siegele, Kompositionsweise, 68.

lsHans Eppstein, "J. S. Bachs Triosonate G-dur (BWV 1039) und ihre Beziehungen
zur Sonate ft’ir Gambe und Cembalo G-dur 03WV 1027)," Musikforschung (1965): 134-
135. Siegele (Kanpositiomweise, 84) suggests that the original was for two recorders
and continuo, in B" Major.

243

Eppstein’s (and Siegele’s) hypothesis is not altogether convincing, however, for
it fails to allow the possibility that Bach may have adjusted parallel passages when first
composing the work in order to accomodate the musical material to the range of the
instrument; such changes need not necessarily betray a copying process from a pre-
existent source. Thus in Example 7-1 above, the octave transposition in the first flute
at m. 12 is fully legitimate within the work itself, and accomodates the range of the flute
to the musical material as presented in the dominant rather than in the tonic as occurs
at m. 24.

On the basis of style and the weight of the surviving sources, Wolff and Dreyfus
maintain that the G major gamba sonata and its parallel trio sonata stem not from Cothen
but Leipzig, and that the only basis for theorizing about an earlier version is the
mistaken notion that Bach’s original chamber works all stem from Cothen, much as the
church compositions are primarily associated with Leipzig. Wolff states:

It seems that BWV 1027 and BWV 1039 are two different means of expressing
the same idea, and that they clearly belong to the same period. The older
hypothesis of a common model is perpetuated primarily by the mistaken idea that
Bach did not write original chamber music whilst in Leipzig, rather only--or at
least primarily-—transcriptions. ‘6
Dreyfus argues for a Leipzig provenance for all three gamba sonatas, citing galant
elements in the second sonata and parallels to the Sonate auf Concertenart in the third.

The amalgamation of sonata and concerto elements in 1029 strongly suggests that the

work originated sometime in Bach’s later years, and definitely in Leipzig.”

 

“Wolff, ”Chamber Music,” 173.

l7Laurence Dreyfus, ed., 3 Sonaten fir Viola da Gamba und Cembalo (Leipzig: C.
F. Peters, 1985), 64-65.

244

Concerto Elements and Source Questions

The third gamba sonata (1029) has no extant counterpart in another version, but
its outward design displays features which strongly resemble a concerto. Spitta noted
that this sonata occupies only three movements like a concerto, and that concerto form
held an important part in the shaping of the allegros.18 The concerto character is most
evident in the first movement, where ritornellos appear throughout: at mm. 1, 11, 35,
64, 73, and 95. Also, in the first two bars of the final ritornello (mm. 95-96), all parts
are in unison, a formula encountered in other concertos such as the Second Brandenburg
(I, mm. 113-114) and the D minor Cembalo Concerto (1052, 1, mm. 172-173 and 184-
190).19 On the basis of these strong concerto-like characteristics, Siegele concluded
that at least the first movement of this sonata is actually a transcription of a (lost)
concerto movement.” In like fashion to Eppstein’s method, Siegele also scrutinized

variant readings of parallel passages within a movement, particularly in the bass as in

 

18Spitta, Bach, II, 118.

19The similar handling of the final ritornello in 1032 and the Second Brandenburg led
Siegele to suggest that the gamba sonata may stem from the same period as the
Brandenburg concerti, i.e. the Cothen years. However, the fact that unison ritornellos
also occur in Leipzig concertos such as 1052 illustrates the weakness of using such
comparisons.

”Siegele, Kompositionsweise, 99. Not all movements display concerto
characteristics equally; Eppstein suggests that the third movement originated from a trio
sonata.

245

Ex. 63, positing that such differences clearly betray the process of transcription from

an earlier work.”

Ex. 6-3. 10291, 7-8, 17-18, 22-23

7-8

 

17-18

 

22—23

 

Closer examination of these passages, however, reveals that in some instances the

alterations serve to avoid problems in voice leading. For example, measures 23 and 8

 

2"'Die ungleichartige Behandlung von Parallelstellen ist . . . Kennzeichen einer
Bearbeitung.” ("The dissimilar handling of parallel passages is a sign of a
transcription." My translation). Siegele, Kompositionsweise, 97. In support of his
hypothesis, Siegele recalls the variant passages in the G major trio/gamba sonatas,
1039/1027 (Ex. 7-1, 2).

246
are parallel passages, the first in G minor, the second in D minor. The reading of the

bass part in m. 23 is altered from that found in m. 8. If the bass pattern of bar 8 were
adopted without change in m. 23 (except for transposition to D minor), parallel octaves

with the gamba would result (Ex. 6-4):

Ex. 6-4. 1029 I, 23 over bass of m. 8

 

Similarly, if the same bass reading from m. 8 were used at m. 18, hidden octaves with

the cembalo discant would occur (Ex. 6-5).

Ex. 6-5. 1029 I, 18 over bass of m. 8

 

In yet another passage not cited by Siegele (m. 12, which parallels m. 2), the use of the
bass from m. 2 with the gamba at m. 12 produces a series of five parallel unisons (Ex.

6-6):

247

Ex. 6—6. 1029 l, 12 with bass of m. 2

 

Such examples suggest that variant readings in the bass may sometimes result from
causes unrelated to transcription, and accordingly do not automatically betray
transcription from a previous work. At the same time, however, it should be stated that
this observation does not negate the possibility of transcription in the case of 1029, but
only questions the basis from which it is presumed. Although conclusive determination
of the early form of this gamba sonata may never be possible, the strong resemblance
to the baroque concerto has led some scholars to speculate about possible models.
Eppstein felt certain that the whole work represented a transcription from one or more
works with two violins, either from a trio sonata, or, in the case of the first and second
movements, perhaps a concerto. The concerto character of this sonata, though
unmistakable, is mitigated somewhat by features which do not follow typical concerto
design, which is perhaps why Eppstein still considered a trio sonata to be a possible
source for this work. A more extreme hypothesis has been offered by Peter Williams
and John Hsu, who have collaborated to produce a transcription of this sonata in
concerto form scored in the manner of the sixth Brandenburg Concerto and now
published by Broude Brothers (1984). Williams conjectures that such a reading might

have been ”in a position to have been chosen as the last of the Six concerts avec

248

plusieurs instruments but that when it was not, its concerto guise dropped from view and
it has come down to us only as a gamba sonata (in a version perhaps belonging to the
Leipzig years). "” This hypothesis particularly demonstrates the extent of speculation
which has been given these works.

Among the other chamber sonatas, the two authentic flute sonatas also display
certain features of the baroque concerto, both in their three-movement form and in
certain details.” The clearest assimilation of such concerto-like features, particularly -
the imprint of the concerto ritornello, is found in both the A major flute sonata (1032)
and the G minor gamba sonata (1029) just discussed. Both works begin with an eight-
measure ritornello which circumscribes the tonality and principal melodic material of the
movement, and follows exactly a type of Vivaldian ritornello sometimes used by Bach.
This ritornello structure is tonally closed and may be conveniently divided into three
parts as follows. The first section defines the tonic by reference to its dominant; the
middle section typically features sequential treatment of motives and avoids reference to

the tonic; and the concluding portion provides a tonal close, ending the ritornello on the

 

22Peter Williams, ”Bach’s G Minor Sonata for viola da gamba and harpsichord BWV
1029: A seventh Brandenburg Concerto?" Early Music 12, 3 (August 1984): 352.

”While the authenticity of the B Minor (1030) and A Major (1032) sonatas is
unquestioned (both survive in autographs), the genuineness of the E" Major Sonata
(1031) has been doubted sufficiently to exclude it from the authentic canon published in
the NBA. Although a most attractive work musically, many scholars from Diirr on have
disqualified the work on stylistic grounds. Dfirr’s conclusion is that it was most likely
composed by Emanuel under Sebastian’s supervision. It is omitted from consideration
in the present study. See Alfred Diirr, ed., Bach, Sonate C-dur fiir Fwte und Basso
continuo BWV 1033, Sonaten Es-dur, g-mollfir Flote und obligates Cembalo BWV 1031,
1020 fiberliq‘ert also Werke Johann Sebastian Bachs (Kassel: Barenreiter, 1975), 2-3.

249

tonic note in the upper voice.” This three-section design is modelled perfectly in the

opening ritornello of the A Major Flute Sonata (Ex. 6-7).

Ex. 6-7. 10321, 1—9

F hate

I raverso

   

Cembalo

 

 

It is notable that this opening eight-bar ritornello is played by cembalo alone, which thus
functions as tutti, immediately followed by the entrance of the flute in the tonic carrying

a solo theme which is not derived from thematic material in the ritornello. The resultant

 

2“Wilhelm Fischer described this three-sectioned ritornello with the terms Vordersatz,
Fortspinnung, and Epilog. Wilhelm Fischer, "Zur Entwicklung der Wiener klassischen
Stils," Studien zur Musikwissenschaft, III (1915): 53. See also Laurence Dreyfus, ".l.
S. Bach’s Concerto Ritornellos and the Question of Invention," The Musical Quarterly
LXXI (1985): 327-58.

250

contrast of thematic material, as well as the change of instrumentation and the three-
section structure of the ritornello, all suggest a concerto. A similar outline may be

observed in the opening of the first movement of the G Minor Gamba Sonata (Ex. 6-8):

Ex. 6-8. 10291, 1-9

I. Vivace

   

Viola da gamba

Cembalo

 

251

There are striking parallels between these two works (1032 and 1029) in the opening
ritornellos; even the length of the ritornellos is the same. However, whereas the
ritornello is entrusted to the cembalo in the flute sonata, in the gamba sonata it is the
gamba which plays the ritornello material over continuo bass. The absence of a cembalo
discant is interesting on several grounds. First, it is clear from the figures appearing
under the bass that when the discant is tacet, the harpsichordist reverts, as it were, to
the role of continuo player.” Second, the eventual entrance of the cembalo discant in
measure 9 assumes an importance similar to that of a first solo entry in a concerto, thus
briefly elevating thelrole of the cembalo above the gamba. Were this a true concerto,
this solo entry would unequivocally announce and identify the solo instrument. And
third, the absence of the ”solo" voice during the ritornello (the gamba taking the place
of the ”tutti") is reminiscent of the pattern employed in the cantatas with obbligato organ
that parallel cembalo concertos 1052 and 1053 (Cantatas 146 and 169 respectively):
during the opening tutti ritornello, the organ discant is tacet, the bass being figured until
the first solo entrance, when both parts are fully written out. The organ thus functions
as continuo during the ritornello and as solo during the solo episodes, much as the
cembalo is called upon to do in this gamba sonata. Among the sonatas with obbligato
cembalo, this gamba sonata is the only one in which the cembalo implies a solo role by
means of a delayed entrance of the discant. In a sense, this design more fully realizes

the title commonly given to these sonatas: Sonata a Cembalo e Viola da Gamba di

 

”The presence of figures under the bass is not unique to this work, as short segments
with figured bass are also found in several of the violin sonatas (1014-1019) and the A
major flute sonata. The inclusion of figures also is apropos to the question, considered
later, whether a second cembalo played continuo in these sonatas.

252

J. 5.3. , rather than Sonata for Viola da gamba (or flute or violin) and Cembalo. Penzel’s
copies of 1028 and 1029 do use the latter reading, but the title page of the autograph
parts for 1027 (Mus. ms. Bach P 226) as well as the autograph scores of flute sonatas
1030 and 1032 (Mus. ms. Bach P 975 and Mus. ms. Bach P 612) all list the cembalo
first in order.

Although the opening "tutti" ritornellos in these two sonatas suggest the form of
the concerto, there are also significant departures from Bach’s usual concerto practice.
Most notable is the lack of true thematic differentiation between tutti and solo material.
In the flute sonata (1032), the theme presented by the flute at its first entrance (mm. 9-
10) gives an initial impression of thematic substance, particularly as the preceding ”tutti"
arrives at a clear cadence on the tonic before the flute enters supported only by continuo
(see Ex. 6-7). But after only two bars of this "solo," the first phrase of the ritornello
returns unexpectedly again in the tonic (cembalo, m. 11), in essence abruptly cutting the
solo episode short even though the flute part continues through a second and third
phrase, the latter again supported only by continuo accompaniment. At the conclusion
of this flute episode (m. 16), the ritornello reenters again for a third time in the tonic

(Ex. 6-9).

253

Ex. 6-9. 1032 1, 8-16

 

Of significance here is the fact that the flute’s opening theme, if it can be termed a
theme at all, never really reappears in the rest of the movement. Thus the flute melody
of mm. 9 ff. assumes no lasting thematic importance, the opening theme of the ritornello.
(mm. 1-3) dominating the movement in both flute and cembalo and even finding its way
briefly into the cembalo bass at mm. 4647, an unusual circumstance in concertos, fugal

movements excepted.”

 

2“A notable exception is found in the first movement of the F major harpsichord
concerto, 1057, where the original violino principale part from the Brandenburg IV
model is transposed down an octave and placed in the cembalo bass several times.

254

The G minor gamba sonata (1029) exhibits many of the same features. The first
entry of the cembalo discant is non-thematic, functioning simply as a counterpoint to the
return of the ritomello’s first two measures, which in turn is followed immediately (m.
11 f.) by a complete restatement of the ritornello again in the tonic key. Further, bars
11-12 are nothing more than an exchange of musical material from the preceding two

bars (mm. 9-10), i.e., Stimmtausch with invertible counterpoint (mm. 11-12) (Ex. 6-10).

Ex. 6-10. 10291, 8-19

I
-'-u——— __ -- _—
: 4- .4" -
'"———-“-4 -.--—w--———---—---
—‘o-— .4 .‘—-——-—-———
——_-—

_.' -,__ , -2:
-----‘ r.- ..-- — ._--* -- mm -
—».____—4 ...——_- .—..-.-. -'_. =.-

.— _

 

255

Thus appear three statements (complete or partial) of the ritornello, all in the tonic and
without any intervening solo episodes, a design obviously not characteristic of the
concerto. Such repetitive use of thematic material, however, is not far removed from
the imitative or fugal manner of trio sonatas. Thus while the melodic and harmonic
design of the ritornello points to concerto, the reiteration of melodic material coupled
with the absence of solo episodes suggests sonata.

This intermixing of concerto and sonata elements has led Dreyfus27 and
Marissen” to suggest that these works represent a genre termed Sonate auf
Concertenart by Johann Adolphe Scheibe.” Scheibe’s definition is not overly specific,
but seems to suggest that the difference between the standard trio sonata and the sonata
in concerto style is largely a matter of counterpoint rather than set formal design. While
it is not difficult to accept the concerto- and sonata-like aspects of these movements, one
must be cautious about accepting Dreyfus’ view that Bach intentionally wrote these

pieces in such a manner as to deliberately confuse the identity of the genre.” This

 

27Laurence Dreyfus, ”I. S. Bach and the Status of Genre,” The Journal of
Musicology, V (1987): 55-78; and Dreyfus, 3 Sonaten, 63.

”Michael Marissen, "A Trio in C major for recorder, violin and continuo by J.S.
Bach?” Early Music Vol. 13 no. 3 (1985): 384-390; and Michael Marissen, "A Critical
Reappraisal of J. S. Bach’s A Major Flute Sonata,” The Journal of Musicology VI
(1988): 367-386.

2“’Johann Adolphe Scheibe, Der Critischer Musikus: Neue, vermehrte und verbesserte
Auflage (Leipzig: Breitkopf, 1745), 677-78.

30Concerning 1029: "The aim here tallies with what appears to be the generic
intention: by encouraging the momentary belief -- here and elsewhere -- that a mere
sonata can assimilate the entire range of codes and devices proper to the concerto grosso,
Bach has thrown doubt on contemporary notions of genre predicated on a common-sense
pecking order. Whereas Scheibe had proposed that a reduction of the concerto into a
sonata was plausible given certain accommodations, Bach questions the logic of

256

leads Dreyfus to the equally unconvincing hypothesis that the resultant ”struggle"
between concerto and sonata genres in the A Major Flute Sonata led Bach to remove
forty-six bars from the autograph of the first movement and rewrite them into a form
which is now lost.31 However, it is also this hypothesis of Bach’s espousal of the
Sonate auf Concertenart that leads Dreyfus to tie these concerto-patterned sonatas, and

their trio models if there were such, to Bach’s later period, i.e., Leipzig.

The Autograph of 1062/1032: A Unique Source

A digression to explan the autograph manuscript of 1032 and its peculiar
condition, alluded to above, is necessary here. This manuscript (Mus. ms. Bach P 612)
was available to Wilhelm Rust at the time of preparation of the BG edition but was lost
during the Second World War, only to be rediscovered and returned to the Deutsche
Staatsbibliothek in 1977,32 more than ten years after the preparation of the flute sonatas
for the NBA.” The autograph P 612 is unique in that it is the only known example of

a double manuscript by Bach in which two separate works of different origin and

 

constraints per se. In so doing, he proposes a formal solution that expands the generic
horizons of both concerto and sonata while endowing the new genre with a special, if
inimitable, identity." Dreyfus, ”Bach and the Status of Genre," 63.

3' Ibid., 63, fn. 12.
32Alfred Dt’irr, NBA KB, VI/3 Supplement, 4.

33The NBA edition of the flute sonatas (NBA Vl/3) appeared in 1963, and Eppstein’s
”Studien fiber J S Bachs Sonaten" also was written and published long before the
manuscript resurfaced. Thus Eppstein’s observations on the A major sonata were made
without benefit of the autograph.

257

instrumentation are presented one above the other in the same manuscript.34 The upper
sixteen staves of folios 1' to 15v contain the score of the Concerto in C Minor for two
harpsichords (1062), while the lowest three staves of the same folios hold the A major
flute sonata. The sonata continues on alone from folio 16r after the concerto leaves off
on folio 15‘.” The concerto has been considered earlier in chapters 2 and 3, but it is
the sonata which draws attention here. The most puzzling feature of P 612 is that the
bottom three staves from folios 9' to 14" (pp. 17-28) have been cut out, with the result
that approximately 46 measures of the first movement are irretrievably lost.” Without
the benefit of actually seeing the autograph, Eppstein suggested that perhaps Bach was
sufficiently dissatisfied with the Vivace as to remove the measures himself, substituting
a revised version which has also disappeared.” This hypothesis has been echoed by
both Dreyfus and Schulze; the latter, with the advantage of the rediscovered manuscript,
also has suggested that the excision may have been the work of Bach himself, since on
pages 17 and 22 the cutting rose above the flute staves and removed a portion of the bass
line above from the harpsichord concerto. The missing bass pitches were then restored

in tablature similar to that used for corrections in the concertos. Schulze claims that

 

34Manuscripts of vocal works in which a solo movement, e.g. an aria, appears across
the bottom staves under a preceding chorus are not unknown in Bach’s oeuvre. But such
an arrangement with two separate, unrelated works from the instrumental repertory is
unique to P 612.

”For a facsimile of the manuscript, see Schulze, ed., Documenta Musicologica.
Zweite Reihe: Handschriften-Faksimiles X (Kassel: Barenreiter, 1980).

”Owing to the lapse of these measures, the BG edition omits the first movement
entirely. The NBA, however, prints the surviving bars but with a footnote at the
appropriate location indicating the absence of ”approximately 46 measures."

37Eppstein, "Studien,” p. 101.

258

these tablature additions are in Bach’s own hand.38 If a second version or revision was
substituted by Bach as suggested by Dreyfus, it has not survived. It seems especially
curious that such a substitute reworking of the excised passage, if it in fact existed, was
not affixed to the manuscript, especially since the bottom staves on the pages
immediately preceding and following the excised passage (leaves 8 and 15) were
apparently also cut out initially but then rejoined with a glued-in strip of paper.

With such little evidence and so much tenuous speculation, it is difficult to draw
firm conclusions or to advance the claim that Bach removed the measures in question
because of a personal disaffection with this work or a recognition of its structural
weaknesses. The weaknesses which continue to be cited by scholars in the extant
portions of the movement obviously were not eliminated by removal of the middle 46
bars; it would have made more sense for Bach to have removed the entire movement if
there was such disaffection with the piece. The present study proffers the hypothesis that
the removal of these staves from the middle of the first movement has nothing to do with
the sonata but rather was due to the concerto, a possibility that neither Eppstein, Siegele,
nor Dreyfus appear to have considered. The autograph in question is comprised of 9
large bifolio sheets, each numbered consecutively at the top by Bach. The page on
which the third movement of the concerto begins (folio 9") is in the first bifolio from
which the bottom staves are missing (bifolio 5). The end of the third movement of the
concerto falls in the middle of the 8th bifolio (on folio 15"), which close scrutiny reveals
was cut in half, separating the end of the concerto from the continuation of the sonata

on the following sheet (16‘). That the 8th bifolio was cut apart is verified by the fact

 

3"Schulze, Faksimile X, 9.

259

that only in this bifolio did Bach number the bifolio twice, once at the top of the first
page (15’), and again at the top of the next, now separated, sheet (16'). Thus the third
movement of the concerto was isolated from the rest of the manuscript, and the bottom
staves of these pages were also removed. These conditions strongly suggest that the
third movement was removed for separate performance, and that the unnecessary bottom
staves containing the flute sonata were eliminated so as not to distract the player of the
second cembalo part. It was suggested in Chapter 2 that the concertos for two
harpsichords were in all likelihood performed by Sebastian with one of his sons or
pupils. In such a scenario, it is possible if not likely that the younger player presided
at the second harpsichord. Particularly if said player was a younger student, e. g. , Krebs
or Agricola, the idea of removing the distracting sonata staves makes all the more
sense.39 The bottom of page 16 (folio 8") and top of page 31 (folio 16') are further
marked with "NB," indicating where the removed bifolios could easily be reinserted.
The separated sheets of the 8th bifolio subsequently were reattached with a strip of
paper, visible on the inside fold of bifolio 8 (the only such strip on an inside fold of any
bifolio), and the bottom staves containing the flute sonata were reattached to folio 15
(pp. 29-30), no doubt after the performance which required the removal of the concerto
movement in the first place. This reattachment thus preserved intact the entirety of the
second movement of the flute sonata which begins on 15". This hypothesis appears to

agree substantially with all the conditions of the manuscript, and also eliminates

 

39Schulze postulated the cembalists for a performance of this concerto at the
collegium concerts: "one would first of all think of Johann Sebastian Bach himself, and
a pupil who frequently officiated at the harpsichord, Johann Ludwig Krebs. " In 1736,
the date of the autograph, Krebs would have been 23 years old. Ibid., 18.

260

speculations about the composer’s disaffection with the sonata. It does not answer why
some parts of the sonata (e. g., the beginning of the second movement) were reinstated
while other measures were not; perhaps it was Bach’s intent to reattach all excised
measures, but some became lost, or perhaps Bach already had another copy of the first
movement of 1032, making restitution of the flute score unnecessary. In any case, if the
foregoing premise is correct, it also suggests that these works, both sonata and concerto,
may occasionally have been played or performed piecemeal, i.e., an isolated movement
by itself, rather than as a complete work. Or perhaps there was a singular occasion
which invited the performance of just this one movement of the C minor concerto."o

Apart from the curious physical condition and layout of P 612, there are some
additional aspects of this manuscript which suggest that the A major flute sonata may
have been derived from some earlier version. First, the text of i the autograph is
remarkably clean, showing no evidence of elective corrections or changes to improve the
musical substance; in other words, this autograph is not a working copy. There are
changes evident in the concerto written above the sonata, but not in the sonata itself,
implying that the sonata was copied from some other source or sources. The only
exceptions to this clean copy status are mistakes which appear to be copying errors, as
in the flute part at I, 22-24, where a passage has been completely crossed out and

replaced with rests. In all likelihood this represents a copying error, for the flute part

 

4oSchulze further speculated about possible performances of this double concerto in
Dresden with his son Wilhelm Friedemann, who became organist at the St. Sophien
church in Dresden in 1733. Such a performance would have coincided with Bach’s
appeals for a court position in Dresden in the midst of his squabbles with the Leipzig
authorities. See Schulze, Faksimile X, 18.

261

in mm. 21 and 27 is identical and the crossed out text (mm. 22-24) reads the same as

mm. 27-29 (Ex. 6-11).‘1

Ex. 6—11. 1032 1, mm. 21-29

21—24, P 612

l
,2

t.
l

r

 

21-29, NBA

 

 

 

 

“NBA KB VI/3, 43-44.

262

Such graphological data helps to confirm that the musical text was copied from
some source, but whether that source was a draft or earlier version of the work is not
ascertainable from this data alone. However, further evidence does suggest that 1032
is a transcription. First, the second movement also exists in a contemporary parts copy
from an unknown hand for violin, cello, and unfigured bass, entitled Concerto a Violino,
Violoncello, et Basso del Sigr. J.S. Bach.42 In this reading, the outer movements are
comprised of a C major version of the first and third movements of the organ Trio
Sonata in E" major (BWV 525). The connection of this ”Concerto” version with Bach
is uncertain, but it is curious that the three movements of the flute sonata all have A as
their tonic (A major-A minor-A major), unique in the chamber works of Bach, while the
key scheme for the ”Concerto" is C major-A minor-C major, a common prototype in
Bach’s chamber literature.“3 Further, the third movement in the autograph betrays
signs of having been copied from a C major source, demonstrated by several copy
mistakes where the pitch was first entered a third too high (Ex. 6-12). Unlike the many
alterations found in the concertos, these changes are minor and are related to the

mechanics of copying the score, not to musical modifications.

 

42Quelle B in NBA KB VI/3, 44.

43It is noteworthy that Eppstein made a case for the key of C Major in the outer
movements on the basis of range and the existence of this ”Concerto" version, but
without knowing about the corrections in the third movement of the autograph. See also
Alfred Dt’irr, Ergc’inzung zum Kritischen Bericht Sonata A-dur fiir Flauto T raverso und
Cembalo, BWV 1032 (Kassel: Barenreiter, 1981), 12-14.

263

Ex. 6—12. 1032 III, 120-122 (P 612)

 

It will also be recalled that the first movement of 1032 strongly suggests the
character of a concerto, complete with a three-sectioned opening ritornello. For all these
reasons, it appears very possible if not probable that 1032 was derived from an earlier
work. Given the strong concerto imprint of the first movement and the lack of same in
the finale, coupled with the disparate works represented in the "Concerto" version, it
may well be that 1032 stems from not one but two or more different models. If this
hypothesis is correct, it may help explain Bach’s seeming carelessness in losing much
of the first movement of the flute sonata when the third movement of the concerto was
removed temporarily from the autograph; i.e., the three movements of the sonata were

not a unity in the first place.
Chronological Ordering of the Sonatas

In attempting to establish a chronological framework for these works, Eppstein
postulated an ordering to all the sonatas with obbligato cembalo based on the structure
and design of the sonata movements themselves. Noting that certain movements are
close to trio-sonata design, others patterned after concerto allegros, and still others

incorporate fugal procedures but with suggestions of tutti-ritornello patterns, Eppstein

264

concluded that such differences indicate a chronological ordering, that is, a progressive
sequence of compositional development. As the violin sonatas (at least the first five)
appear to share more advanced features such as tutti-fugues and accompanied solo-type
slow movements, Eppstein found in these the final stage of Bach’s thinking about the
sonata.“ Consequently, the flute and gamba sonatas, or at least their first models,
were dated earlier, with the flute sonatas as a group preceding the gamba sonatas. While
this ordering may have some merit, it does treat the issues somewhat artificially,
implying that Bach wrote only one kind of sonata for a particular instrumentation at a
time. Only after finishing work on flute sonatas did he then turn his attention to gamba
sonatas, and so on.

While style and textual evidence is certainly important, it must not overshadow
the weight of source evidence. Most of the surviving sources for the sonatas with
obbligato cembalo (and their models/variant versions), whether autograph or copies, are
datable by watermark to the 1730’s or later. The following table (Table 7-1) lists the

relevant sonatas with the dates associated with their papers."5

 

“Hans Eppstein, "Grundziige in J. S. Bachs Sonatenschaffen," BJ 55 (1969): 12.
”Watermark dates drawn from NBA KB lX/2.

265

Work Watermflk dates
Gamba Sonatas
BWV 1027 [P 226] (autograph) May 1739-174246
BWV 1039 (trio sonata) [St 431] perhaps 1735-45
BWV 1028 [P 1057] Penzel copy dated 1753
BWV 1029 [St 163] Penzel copy dated 1753
Flute Sonatas
BWV 1030 [P 975] (autograph) 1736-49
BWV 1032 [P 612] (autograph) 1732—34"7
Violin Sonatas
BWV 1021 [60.8.3] 1732-34 (from dated MS)
BWV 1038 (trio sonata) 1732-3448
BWV 1014-1019a [St 162] used 1712-1729; in Leipzig 1723

(partial autograph)

Table 6-1. Watermark dating of obbligato sonatas.

What is most interesting about this list is the observation that the works with the
earliest datable source are the violin sonatas, the very works with the most ”advanced"
features. Since there is evidence to support the existence of the violin sonatas at the very
beginning of Bach’s Leipzig tenure (0. 1723-24) if not even earlier, Eppstein was forced

to backdate the flute and gamba sonatas to Cothen, since he postulated a developmental

 

‘6 Kobayashi suggests c. 1742; Eppstein in his notes to the KB of the NBA (VI/4)
places the appearance of the autograph ”in the 1740’s.” See Kobayashi, "Zur
Chronologie," 50; NBA KB VI/4, 11.

4“’For the G minor version of BWV 1030, Marshall suggests 1729, ”perhaps in
connection with Bach’s having assumed the directorship of the collegium musicum, " and
1736 for both BWV 1030 (now in B minor) and BWV 1032, ”perhaps, again, with a
view toward a performance with the collegium musicum in Leipzig--or, conceivably by
Buffardin in Dresden." Marshall, "Flute," 496.

“It is especially notable that the related pair of works BWV 1021 and 1038 both
appear on paper bearing the same watermark (large MA), which watermark is also
shared by the A major Flute Sonata, BWV 1032.

266

hierarchy for these sonatas. However, if all of these sonatas really date to Cothen, as
most writers since Forkel have suggested, then why are there no extant source materials
from Cothen, either scores or performing parts? And if the above sources from Leipzig
do represent only copies of versions composed earlier, why is it that the copies survived
but none of the originals, even though the originals obviously had to have been extant
when the copies were penned? Admittedly, the division of Bach’s estate among his heirs
had a direct effect on the preservation of these materials, as some members of the Bach
family circle were more careful than others in safekeeping manuscripts entrusted to
them. The vocal works, considered outmoded by 1750, gained little continuing attention,
but the instrumental works may well have received performances after Sebastian’s death,
which may have contributed to the loss of some sources and even whole works.49 But
source questions notwithstanding, there are also stylistic anomalies if these works are
associated with Cothen. The largo of the B minor flute sonata (1030) is close to the
accompanied-solo type movement, and, as Wolff points out, the expansive design of the
first movement as well as the uniqueness of the overall structure of the sonata has no
parallel with any pre-Leipzig instrumental work."0 There also exists a copy of the
clavier part only for 1030 in a version in G minor which is generally regarded as being

earlier than the B minor version.51 Marshall posits a date between 1729-31 even for

 

49For this observation I am indebted to Christoph Wolff, who shared this thought
during a conversation on related questions.

”Wolff, ”Chamber Music," 173.

51Mus. ms. Bach P 1008 in the Westdeutschen Bibliothek, Marburg. This source is
Quelle D in NBA KB Vl/3, 32, and is reproduced in the appendix to the Bericht, pp. 89-
103. It is impossible to establish an exact date for this source, since it survives only in
a posthumous copy from the later 18th Century.

267

the earlier G minor version, thus placing both versions in Leipzig during the Collegium
years.’2

Unlike the flute and gamba sonatas, the violin sonatas (1014-1019) seem to
comprise a closed group; in all seven sources cited in the NBA, the six sonatas appear
together in the same order, suggesting that they have long been considered as a unit.
The first five sonatas all consist of four movements each in the traditional ordering of
the sonata do chiesa, but Sonata 6 in G Major is exceptional in the number and structure
of its movements. In the absence of an autograph, it is impossible to speculate whether
Bach conceived these sonatas, especially numbers 1-5, as a unified group. Forkel
asserted that all six sonatas derived from Cothen, to which Spitta added that this
attribution ”must have emanated from Bach’s son. "53 Most modern scholars have
accepted Cbthen as the place of origin, including the editors of the NBA who place the
sonatas c. 1718-22.54

Of the seven source manuscripts listed in the NBA,” the most reliable was

copied by Bach’s pupil and son-in-law Joh. Ch. Altnikol sometime between 1744 and

 

”Marshall, 485. Marshall bases his dating of the G minor version largely on a
resemblance to the opening chorus of Cantata 117, Sei Lob und Ehr’ dem Mchsten Gut,
composed between 1728 and 1731. This argument is tenuous, however, since even if
the sonata theme were indebted to the cantata, the time of composition of the two works
need not coincide. Marshall also questions the existence of earlier lost versions of both
flute sonatas, thus assigning both works to Leipzig.

53Spitta, Bach, Vol. II, p. 110. Spitta also adds that these six sonatas were ”united
by Bach into a whole set," and that the association with Cothen comes from "a very
credible tradition." The ”tradition" referred to is none other than Forkel’s attribution.

“Giinter Hausswald and Rudolf Gerber, NBA KB VI/ 1, 141.

5"NBA KB VI/l, 137-141.

268

1758, and presents the sonatas in their latest, final form. This manuscript (Mus. Ms.
Bach P 229) may have been copied as early as 1744 during Bach’s last years, or as late
as 1758 (Altnikol died in 1759).56 Excepting revisions in the cembalo part of the third
movement of 1018, the first five sonatas are all without essential changes or differences
between the various sources. The sixth sonata (1019), however, exists in three strikingly
different versions which present a complex array of movements. Unlike the previous
five sonatas, each version of this sonata contains five or six movements, only some of
which are common to all three versions. Table 7-2 ennumerates the content of each

version (lower case letters identify common movements).57

 

56Altnikol became one of Bach’s pupils in 1744. The title of P 229 identifies the
copyist as Bach’s ”son-in-law and student the Naumburg organist Altnikol. " As this title
is in a later hand, the date could be as early as 1744, before Altnikol actually assumed
the organist position in 1748. Gerber and Dfirr place the copy not earlier than about
1747. See NBA KB, VI/1, 138; Alfred Dfirr, "Zur Chronologie der Handschrift Johann
Christoph Altnikols und Johann Friedrich Agricolas," BI 56 (1970): 46-48; and Hans
Eppstein, 'Zur Problematik von J. S. Bachs Sonate ffir Violine und Cembalo G-dur
(BWV 1019)," Archiv fiir Musikwissenschaft xxi (1964): 223.

”The ordering of the versions is adopted from the NBA, which also lists the various
movements for the three versions. See NBA KB VI/ 1, 201.

269

 

Version I II III
NBA Source: C,D E A,B,F,G
a) Presto, G maj. a) Vivace a) Allegro
b) Largo, E min. b) Largo b) Largo

c) Cantabile ma un
poco adagio, G maj.
e) cembalo solo, E min.
(=Courante)
g) cembalo solo,
E min.
d) Adagio, B min. d) Adagio
h) Adagio, B min.
f) Violin solo
w/bass accomp., G min.
(=Gavotte)
a) Presto a) Vivace
i) Allegro, G maj.

Table 7-2. Variant versions of Violin Sonata, 1019.

From an examination of different readings of the bass in the first 11 measures of the
Largo, Eppstein has argued convincingly that source E represents the earliest version of
1019.58 In this source, the bass in bars 1-11 is notated in even quarter notes, while in
sources C and D the same passage is filled in with eighth notes, producing a much more

florid line (Ex. 6-13).

 

58Eppstein, ”Problematik,” 226-228.

270
Ex. 6—13. 1019, 11, mm. 1-11, sources C,D over bass of E

Largo

    

 

 

 

 

“41 - J . fl—fi
I [IV ‘ I ll - IV I
[LL-K 1.: I {Er y l I l I [I l l :- l l I
[k l I I I II ' .1 -
‘Vfil
I; I I I ‘r l v :} fl
1

 

 

 

 

 

 

[F ; : : h- r I A [I fl
_ 1 ‘1? l— - r I Ifi
T :1 J I‘- r 'I [IV I r ‘1 5‘ fl I
g ' v I l I: 1 l l - ‘ j

 

 

.-
4 1-
1 4L
1 I

W71]
i

It is unlikely that Bach would have written a simpler version of the passage after having
already composed it in a more advanced form (only to return to the more advanced
reading later in source A). The more fluid bass, coupled with several ornaments added
into sources A, C, and D, all suggest that source E is the earliest.

This observation is of some consequence. Substitution of smaller note values in

the bass illustrates a process of embellishment commonly used by Bach when

271

transcribing such works; examples have already been seen in the concertos, and will be
sought in the sonatas in the next chapter. More importantly, however, source E (Mus.
ms. Bach St 162) is a partial autograph; Bach’s own hand is visible in a portion of the
keyboard part in movements 3-5.59 What both Gerber and Eppstein failed to consider
was the identity of the copyist of the remainer.“o It has since been discovered that the
keyboard part in source E is primarily in the hand of "Hauptkopist C,” an important
scribe represented in manuscripts of Bach’s vocal works between April 1724 and
February 1727."I The dating of source E is narrowed further by the fact that
movements 3 and 5 are early forms of two movements found in the Klavierbu’chlein fiir
Anna Magdalena Bach of 1725, and which later appear as the Corrente and Gavotte of
the E-minor Clavier Partita, BWV 830.62 This means that source E must have been
copied in 1724 or 1725, during Bach’s first years in Leipzig. Thus temporal boundaries
may be established for all the extant sources, and notably they all originate in Leipzig.

It is nonetheless still possible that Bach could have composed the first versions
of these sonatas in Cethen, and then had them recopied after settling in Leipzig. The

intermediate and final versions would then have been composed in Leipzig sometime

 

59NBA KB, VI/ 1, 140. This is the only portion of any of these sources which is an
autograph. A facsimile of the beginning of the third movement, cembalo part, in Bach’s
hand is reproduced in NBA VI/ 1, ix.

”Eppstein concluded that the copyists of sources C, D, and E were "unknown," with
the exception of the small autograph portion in E. Eppstein, "Grundziige," 223.

61Marshall, 475. "Hauptkopist C" has been identified by Kobayashi as Johann
Heinrich Bach. See Yoshitake Kobayashi, ”Zur Chronologie der Spatwerke Johann
Sebastian Bachs Kompositions- und Aufffihrungstatigkeit von 1736 his 1750," B] 74
(1988): 27.

62Marshall, 476.

272

between 1724 and the mid to late 1740’s, since it is unlikely that Bach would have
recopied in Leipzig the earlier versions of a work already brought to final form in
Cothen. Thus it can be concluded with reasonable assurance that the final form of the
sonatas for violin and obbligato harpsichord was realized in Leipzig, not C6then. It must
be remembered, however, that the first five sonatas (excepting revisions in the keyboard
part of 1018, III) are all without essential differences between the various extant sources,
i.e., these sonatas are virtually the same between the early source E of 1724/1725 and
Altnikol’s later copy. Hence the final form of most of these sonatas may be the same
as the first reading. Without a complete autograph it is impossible to know how close
in time source E is to the first composition of these works. If very close, then they may
have originated in Leipzig; otherwise, Cothen may still be possible.

The issues concerning chronology are of considerable significance. If a number
of these chamber sonatas did in fact originate from Leipzig rather than Cothen, then a
totally different picture of Bach’s dealings with this genre emerges. The violin sonatas
are pivotal to this understanding, for they show a stylistically more advanced approach,
particularly with regard to the cembalo. One has the feeling that if it were not for the
early source date of St 162 (source E), scholars might have dated these sonatas later on
the basis of style criteria. Eppstein was forced to place the flute and gamba sonatas
earlier due to the more advanced character of the violin sonatas, since he promoted the
premise that Bach’s compositional activity in the sonata genre gradually went through
a certain metamorphosis by which he eventually realized the accompanied solo type,
freed from the shackles of trio sonata design. In this view, the Musical Ofl'ering is

somewhat anachronistic in its inclusion of a strict trio sonata. Eppstein’s basic premise

273

remains that the four-movement sonata with obbligato cembalo incorporating fugue is on
a higher plane, representative of Bach’s final thoughts on the sonata. If, on the other
hand, the flute sonatas and first gamba sonata were new in Leipzig in the 1730’s, and
thus appeared after the violin sonatas and contemporaneously with the keyboard
concertos, then a significantly altered understanding of Bach’s developmental journey
through the sonata accrues. If the latter hypothesis is correct, then several types of
sonata existed side by side and contemporaneously in Bach’s output, much as two
different approaches to the harpsichord concerto were identified and found to be
coexistent. This is not to deny that some works may be more musically advanced, such
as the B minor and F minor violin sonatas; but it does negate the view of a progressive,
conscious effort on Bach’s part to elevate the chamber sonata. It is notable that in
Leipzig Bach appears to have sought to synthesize various stylistic trends, seen in Book
II of the Well-Tempered Clavier, the B minor Mass, the Musical Ofl’ering, and other
works. It is not difficult to see such amalgamation present in the gamba sonatas and the
A major flute sonata, where concerto and sonata join. It is perhaps too much to state,
as Dreyfus attempts, that Bach deliberately sought to "confuse genres” and write in the
newer Sonate auf Concertenart described by Scheibe. But the premise that Bach
naturally, as part of his instinctive musicianship, synthesized musical characteristics into
one musical whole is not foreign to his musical purpose. In such a view, parallel works
become ”different means of expressing the same idea, "‘3 perhaps for different occasions
or functions or to fit changed performance requirements. This view does not negate the

recognition on stylistic grounds that some of these works look backward while others

 

63Wolff, ”Leipzig Chamber Music,” 173.

274

appear more progressive. To the latter clearly belong several of the violin sonatas, and

it is in these that the cembalo takes on a new and more important role.

Structure and the Cembalo in the Violin Sonatas

Each of the first five violin sonatas (BWV 1014-1018) is in four movements
following the pattern of the sonata da chiesa. Only two other sonatas with obbligato
cembalo, the G major and D major gamba sonatas (BWV 1027 and 1028), follow this
scheme. Despite the outward resemblance to the sonata da chiesa, however, the formal
designs within movements are quite varied. Almost all the allegros are fugal, though
with some distinct differences among them. One group of such fugal allegros, including
1014 IV, 1015 IV, and 1018 II, are in a binary form, with repeat signs clearly marking
the binary division. Of greater interest is a type termed by Eppstein ”tutti-fugue”
modeled after the concerto form of Vivaldi, in which the fugue exposition and reentries
of the theme occupy the place of ritornellos, while passages between the tutti sections
emulate solos.64 Some of these allegro movements also incorporate a da capo repeat
of the opening bars (= exposition), thus suggesting a closing ritornello.” In the A
major sonata (1015), the intimation of tutti material is increased still further by the
appearance of brief passages where all three parts are in unison, as at mm. 29 and 121-2

of the second movement (Ex. 6-14).

 

“Eppstein, "Grundziige," 9.

“This da capo structure is present in 1014 II, 1015 II, 1016 IV, and 1019 V.

275

Ex. 6-14. 1015 II, 28-30

 

Reminiscent of similar segments in some concertos, these passages set off the following
"solo” episode very clearly or highlight final cadences. Both these functions are most
clearly illustrated in the D minor cembalo concerto (1052), where the opening tutti is
played in unison up to the first entrance of the solo cembalo at m. 7. This same unison
tutti material returns (Bach simply writes ”da capo" at m. 185 in the autograph) to
conclude the movement at mm. 184-190.“6

In some respects, it is the slow movements of the violin sonatas which provide
the greatest interest and evidence of stylistic advancement. Two approaches may be
distinguished: one which is closest to the trio sonata, and the other which features a
melody instrument in an accompanied solo. The essence of the baroque trio sonata lies
in its contrapuntal texture, whereby three voices are imitatively maneuvered in such a
way that each is distinguishable but none assumes priority over the others. This
traditional pattern of the trio sonata is clearly evident in slow movements of the A major

and G major violin sonatas (1015 and 1019 respectively). Typically, the upper voices

 

“Further use of such unison tutti passages in a concerto may be found in the third
Brandenburg, 1, mm. 7, 57, 125, and 135-6. Measures 135-6 occurs at the very end of
the movement, similar to the unison cadential passage in bars 121-122 of the A major
violin sonata or the concluding bars of the D minor cembalo concerto.

276

form an equal-voiced duet, usually characterized by imitation, with the bass sometimes
participating in the imitation as well. The kinship with trio sonata design is thus ‘
unmistakable. The beginning of the first movement of the A major

sonata (1015) illustrates these characteristic features (Ex. 6-15).

Ex. 6-15. 1015 II, 1-6

 

In the third movement of the same work, the bass does not participate in the imitation
but the upper two voices are in strict canon at the unison throughout, demonstrating a
degree of contrapuntal control not encountered elsewhere in these sonatas. Along with
the two slow movements of the G major sonata (1019), these movements look backward
to the traditions of the well-established baroque trio sonata.

The second type of slow movement moves in a different direction, the cembalo
and solo instrument being largely independent of each other thematically but mutually
dependent musically as solo and accompaniment. This type is related to the lyric aria

or concerto slow movement: an expressive cantabile melody over a self-sufficient

277

accompaniment which deliberately develops a figure or motive, sometimes over an
ostinato or quasi-ostinato bass. Such melodic motifs are often quite idiomatic to
keyboard, and represent a real departure from the typical trio sonata. Examples may be
found in all the slow movements of the B minor, C minor, and F minor sonatas (1014,
1017, and 1018), as well as the first movement of the E major sonata (1016). There is
much variety in these movements, but what is particularly striking is the greater place
given the cembalo. In the absence of imitation between solo and keyboard instruments,
the latter gains a degree of independence not found elsewhere. This relationship is
particularly noticeable in the opening movements of the B minor and F minor sonatas
(BWV 1014 and 1018). In the first of these, both the discant and bass of the cembalo
function independently of the violin part, until in the second half of the movement the
violin takes up the thematic motive which has been the property of the clavier for 15

bars (Ex. 6-16).

278

Ex. 6—16. 10141, mm. 1-6, 16-20

Violino

Cembalo

 

 

 

 

It is notable that the cembalo is in three parts throughout, even when the violin has
double notes beginning at m. 16. Strictly speaking, only three parts are possible in a trio

sonata texture. The additional voice in the cembalo discant in this movement is not only

279

more idiomatic to keyboard, but also signals the degree of departure from typical trio
practice.67 At certain points (e.g. mm. 11-15, 24-30), the two parts in the discant
separate and show some degree of contrapuntal independence. The arpeggiated figure
in the bass is heard regularly enough to suggest an ostinato.

Many of these features are also present in the opening movement of the F minor
violin sonata (1018), the cembalo again having not two but three voices, this time
organized around imitation of a short motive which comprises the principal
melodic/thematic material of the movement. By contrast, the melodic material of the
violin part is completely unrelated and much less interesting than what appears in the

cembalo (Ex. 6- 17).

 

67The question might be asked whether Bach’s placement of this sonata with its
multi-part opening movement at the very outset of these six violin sonatas was not a
deliberate stroke to mark the departure of these works from the customary sonata genres.

Ex. 6-17. 10181, 1-19

__——— .
—
__._— _- ,— —-__.————
{$L__-- 7:: — _. .7..,.__————_—..—_ ,‘

 

It is noteworthy that the long values and extended tacet passages of the violin in essence
promote the cembalo to the role of principal instrument, the long—breathed lines of the
violin suggesting more an ornamental accompaniment than a solo part. Not to be
overlooked is the sheer length of this Largo: 106 bars, by far the mom expansive of any
slow movement in the chamber sonatas. Though shorter, the third movement of the
same sonata is also quite original, and in effect corroborates the new relationship

between keyboard and solo instruments. In this case, the violin has nothing but doubled

281

eighth notes without melodic importance while the cembalo discant is marked by a
recurrent motive in 32nds. What results is a total role reversal: the violin, normally the
solo instrument, provides the harmonic outline (formerly the role of the harpsichord”

continuo) while the cembalo assumes the character of a solo part (Ex. 6—18).

Ex. 6—18. 1018111, 1-5

Adagio

 

282

These movements (1018 I, III and 1014 I) clearly represent a new relationship between
solo and cembalo; Eppstein termed it "Bach’s most bold advance into new chamber
music territory."68

The third movement of 1018 cited above exists in two versions, one from
Altnickol’s copy of the 1740’s (NBA source ’A’), the other from earlier (NBA source
’C’). The latter presents the cembalo part in a much simpler form, right and left hand

parts both being in open arpeggiated sixteenths (Ex. 6-19).

Ex. 6-19. 1018 111, 1-7, source C

 

Although the harmonic outline is the same, the two readings are worlds apart. Not only

is the later version more idiomatic to the keyboard, but the parts are- now distinguished

 

68"Diese beiden Sitze stellen nach Satzstruktur und musikalischem Charakter einen
der kfihnsten Vorstosse Bachs in kammermusikalisches Neuland dar.” My translation.
Eppstein, ”Studien,” 43.

283
from each other: the arpeggiated bass outlines the harmony as before (though in 32nds

which match the right hand rhythmically) while the scalar discant transforms what had
been simple arpeggiated triads into a melodic motive, the identity of which is highlighted
by always beginning on a dissonance that resolves down by step. Unfortunately, the
violin part of the earlier version (source C) does not survive, so a complete
reconstruction of the earlier reading showing the relationship between the two
instruments is impossible. It is quite plausible, however, that the violin part in the
earlier version might have been different, since the violin reading of source A played
with the cembalo part of source C would create a very austere result. In effect, this
means that the more florid cembalo reading of source A makes the open intervals of the
violin part possible, i.e., that the violin part is controlled by the cembalo figuration.
Thus while the three parts are thematically independent, there is a musical
interdependence which validates the whole.

The sixth sonata in G major (1019) is in many respects unlike any of the other
violin sonatas. That it begins with an allegro movement rather than the customary
adagio is but one indication of this difference. While the slow movements, as already
noted, are closer to the older type trio sonata, the allegros move beyond the confines of
the typical baroque sonata. The first movement particularly reveals the hybrid nature
of this piece: the character of the opening four-bar theme strongly suggests a concerto,
while the restatement of the same theme on the dominant five bars later in the cembalo
discant implies fugue. However, the opening theme never appears in the bass, nor does
it reappear anywhere else in the movement until the da capo restatement of the first 20

measures beginning at m. 70. Between these outer A sections (mm. 1-21 and 70-91) is

284

a B section without real solo character but which is handled fugally, the theme even
appearing in the bass (mm. 33 ff.). Thus what results is an amazing synthesis of
elements from concerto, fugue, and sonata all brought together in one movement.
Eppstein calls this movement a "Concerto allegro with da capo,” a somewhat puzzling
designation since most of Bach’s concerto allegro movements include a da capo for the
closing ritornello; in the autographs of almost all of the harpsichord concertos Bach
simply writes ”da capo" at the appropriate place for the final ritornello.

The third movement of 1019 (in the final version) is also unique in its
instrumentation for cembalo alone. Though employing three part-trio texture (especially
in the first half), the movement shows the imprint of the suite: binary formal design, the
second half commencing on the relative major of the key with the musical material of
the opening. At times the keyboard writing is quite brilliant. The finale, also newly
composed for this later version, is another tutti-fugue in which the theme is heard in all

voices, including the bass.
The Second Keyboard Question Revisited

Lastly, passing consideration should be given to a question which arose earlier
in the cembalo concertos: whether a second harpsichord was expected to play a continuo
bass part. The typical baroque sonata, whether trio or solo, assumed the participation
and support of the continuo. The question arises whether this expectation was eliminated
by the substitution of an obbligato cembalo in the small chamber sonata genre. Ruth
Halle Rowen, citing an article by Hans Joachim Moser, states that Kimberger had made

a copy of the six violin sonatas, termed "Trios" with Fundamento by Kirnberger, which

285

called for two keyboard instruments."9 Apparently Rowen has misunderstood Moser’s
essay, for nowhere in his article does Moser refer to the simultaneous use of two
harpsichords.70 The manuscript Moser refers to is no doubt the NBA’s source E, which
bears the exact same title including the strange spelling of "Sonata": Sei Sounate a
Cembalo certato e violino solo, col Basso per Viola da gamba accompagnato, se piace
composte da Giov: Sebast: Bach.71 To further clarify the identify of the source, Moser
mentions that this manuscript is a partial autograph which once was in the possession of
Franz Hauser. In all points, this description matches that in the NBA except that the
latter makes no mention of the manuscript’s provenance from Kirnberger. It is
interesting to note that Moser in 1938 identified this source as the first version of the
violin sonatas, inasmuch as Hausswald and Gerber in the NBA (1958) believed it
represented the second reading.72 The title, reproduced above, clarifies itself: "col
Basso per Viola da gamba accompagnato, se piace," that is, with the added
accompaniment of a gamba (not second harpsichord) if desired. Moser interprets this
optional strengthening of the bass to mean that if a keyboardist is using a harpsichord

on which the bass is too weak in sound to match the strength of the violin, then a

 

69Rowen, Chamber Music, 126.

70Hans Joachim Moser, ”J.S. Bachs sechs Sonaten fiir Cembalo und Violine,"
Zeitschrzft fur Musik, cv (1938): 1221.

71Beneath the title and in another hand from the 18th century is added the curious

(and obvious) ascription "NB. Diese Trio hat er vor seinem Ende componiert. " NBA
KB VI/l, 139.

72As noted earlier in this chapter, this assessment has been convincingly refuted by
Eppstein, who has reaffirmed source E (Mus ms. Bach St 162) as the first version.

286

gambist might be invited to double the bass.73 Such logic is met with only rarely in
these studies. A further manuscript source listed in the NBA (source G)74 also
mentions this "Basso,” but it is quite evident that the title was derived directly from
source E. It reads: Sounate a Cembalo concertato Violin Solo Basso per Viola da
Gamba accompagn: Se piace Composte da Giov: Sebast: Bach. In virtually every detail,
punctuation included, this title matches the wording of the title from source E, even
though the musical text of source G is the same reading as source A (Altnickol’s copy
of the later version).

It seems apparent that for the sonatas with obbligato cembalo, there was no
assumption on the part of the composer or contemporary performers to add a continuo
part; the cembalo is self-sufficient to supply the thematic and harmonic substance of the
work. In those passages where the cembalo discant is silent, it is very likely that the
harpsichordist may have filled in the harmony, continuo fashion, where indicated by the
bass figures reproduced in brackets in the NBA. According to the critical notes, these
figures are taken from sources B through G.” No doubt given the pervasive common
practice of the time, there were performances of these works even past 1750 in which
performers found it unthinkable to omit the continuo bass, even if only a gamba. The
lack of any textual evidence to support added continuo in these sonatas may also
strengthen the conjecture that such a practice similarly was not mandated for the

harpsichord concertos either. The use of a basso continuo line is preserved even well

 

73Moser, ”Sechs Sonaten," 1221.
74Mus. ms. Bach St 403. See NBA KB VI/l, 141.
”Ibid., 143.

287

into Mozart’s piano concertos, suggesting that the practice did not die quickly nor easily.
As mentioned earlier, in the absence of autograph parts for all the harpsichord concertos,
the existence of a continuo part for 1055 does not provide conclusive proof on this issue
as it relates to the concertos. But the suggestion in the sonatas that a continuo bass may
be added optionally may give some further insight into the contemporaneous practice vis—

a-vis continuo use in all instrumental chamber music.

VII.

The Obbligato Cembalo Part in the Sonatas

The previous chapter was devoted to questions of dating and the formal designs
of the various sonatas with obbligato cembalo. This final chapter considers Bach’s
handling of the cembalo part itself, with particular attention to two issues: Bach’s manner
of writing for an obbligato harpsichord (as distinct from continuo) in a sonata; and
comparisons with the cembalo writing in the harpsichord concertos.

For the sonatas, the methods of inquiry are necessarily more limited than those
applied to the concertos. First and foremost, the sonatas lack the significant number of
extant models or parallel versions which exist for the concertos. With but one exception,
this circumstance eliminates the model/transcription comparisons that illuminated Bach’s
compositional process in the concertos. Second, the sonatas, unlike the concertos, lack
a separate continuo bass line, thus limiting the freedom of the cembalo bass. Third, the
greater reliance on fugal writing in three-voice texture in a number of the sonatas also
restricts the free handling of the cembalo, especially the discant. And fourth, sonatas
are, by their very nature, less given to outward displays of brilliance than is customary
in a concerto. In the latter, the harpsichord is primary, the soloist; but in a sonata, the
harpsichord at best shares the weight of musical responsibility with the solo instrument.

In all likelihood, the fashioning of an obbligato cembalo part in place of the

continuo was not such a demanding task for Bach, if contemporary accounts of Bach’s

288

289

own thorough-bass playing are to be trusted. Lorenz Mizler, Bach’s former student,
reported in 1738:

Whoever wishes truly to observe what delicacy in thorough bass and very good
accompanying mean need only take the trouble to hear our Capellmeister Bach
here, who accompanies every thorough bass to a solo so that one thinks it is a
piece of concerted music and as if the melody he plays in the right hand were
written beforehand. I can give a living testimony of this since I have heard it
myself.1

The date of Mizler’s report is not without significance. In 1738 Mizler founded the
Society of Musical Sciences (Societat der musicalischen Wissenschaften) which Bach
eventually joined, with apparent reluctance, in 1747 . It was on account of Bach’s entry
to the Society that the now famous Hausmann portrait was painted and the Canonic
Variations composed. Also, 1738 is in the middle of the Collegium years. Mizler’s
reference to hearing Bach’s continuo realizations well may have arisen from that source.
Similar accounts were recorded by others, including individuals outside the Bach
circle. Johann Friedrich Daube, a proponent of Rameau’s theories of harmony in
Germany, wrote in 1756:
When he [Bach] played, the [written out] upper voice had to shine. By his
exceedingly adroit accompaniment he gave it life when it had none. He knew
how to imitate it so cleverly, with either the right hand or the left, and how to
introduce an unexpected counter-theme against it, so that the listener would have
sworn that everything had been conscientiously written out. At the same time,
the regular accompaniment was very little curtailed. In general his accompanying
was always like a concertante part most conscientiously worked out and added as
a companion to the upper voice so that at the appropriate time the upper voice
would shine. This right was even given at times to the bass, without slighting the

upper voice. Suffice it to say that anyone who missed hearing him missed a great
deal.2

 

lThe Bach Reader, 231.
2Ibid., 256.

290

Several of Daube’s observations are particularly illuminating. Especially noteworthy is
the reference to Bach’s method of drawing musical motives from the solo instrument line
and incorporating them imitatively in either right or left hand, i.e. in the discant or the
bass, and that this was done so adroitly as to suggest that the parts had been worked out
on paper ahead of time. Further, Bach’s continuo playing was reportedly "like a
concertante part,” meaning that even though it was extemporized from figured bass, it
sounded like an obbligato part, including soloistic passages in balance with the solo
instrument. At the same time, the ”regular accompaniment" (i.e., continuo) was "little
curtailed: " in other words, both obbligato and continuo accompaniments were managed
simultaneously. It is significant that many of these observations parallel what one finds
in the chamber sonatas with obbligato cembalo: exchange of thematic motives between
parts, including the solo line; filled in continuo harmony when the discant is tacet; and
the periodic presentation of thematic material in the cembalo discant and bass. It is not
inconceivable that the obbligato cembalo writing in some of these sonatas may
approximate Bach’s own continuo realizations.

That Bach’s continuo accompaniment was not limited to simple chordal harmony
but was enriched by thematic detail is suggested also in a letter to Forkel from Emanuel
Bach in the 1770’s. Emanuel remarks that Sebastian "accompanied trios on more than
one occasion on the spur of the moment and . . . on the basis of a sparsely figured
continuo part just set before him, converted them into complete quartets, astounding the
composer of the trios."3 This account implies the extemporizing of an additional

concertante part above the bass, most likely in the discant, thus realizing a total of four

 

31bid. , 277.

291

separate, identifiable parts. Emanuel’s account is not sufficiently detailed to know if
Sebastian also simultaneously supplied the figured harmony; if he did, then his practice

agrees with Daube’s account cited above.

Model/Transcription Comparisons in the Sonatas

Among all of Bach’s sonatas with obbligato cembalo, the G major Gamba Sonata
(1027) is the only work which also exists in another authentic version as a trio sonata
in G major for two flutes and continuo (1039). Eppstein’s hypothesis that many if not
most of Bach’s solo sonatas with obbligato cembalo were written originally as trio
sonatas finds its clearest exemplar in this related pair of works. The traditional view as
held by Rust, Spitta, Schweitzer, and others was that the trio sonata was written first and
the gamba sonata transcribed from it. That the latter is written in strict trio style
throughout would appear to support such a hypothesis. As noted in chapter 6, there is
also a third version of the first, second, and fourth movements of the work, apparently
for pedal clavier or organ. Discrepancies between the three readings has prompted the
hypothesis that both 1027 and 1039 go back to an earlier, now lost, original model.4
Of the three extant versions, the trio (1039) is in all likelihood the earliest and is
especially important for providing the only opportunity to compare a model and
transcription of a sonata analogous to comparisons between paired versions of the

concertos .

 

4See Chapter 6, fn. 13. A detailed discussion of the trio, gamba, and pedal clavier
versions may be found in Eppstein, "Studien," 129, and Siegele, Kompositionsweise, 68.

292

These two sonatas (1027 and 1039) are virtually identical in essentials, with the
three-voiced character of the trio preserved in the gamba version. As compared with the
concertos, there is far less substantive change or alteration of musical material between
trio and gamba versions, but the allocation of parts is notable. The continuo bass of the
trio becomes the bass line for the cembalo in the gamba sonata, and it might be expected
that the second treble from the trio would be transferred to the cembalo discant. Rowen
suggests that this was a common transcription procedure in the period:

The trio sonata itself was an important source from which the solo sonata derived

its distribution of instruments. In this case, the process of transformation was

effected by condensing the two upper lines into one, or by transferring the part

for the second melodic instrument of the trio sonata to the right hand of the

keyboard.5
In this G major gamba sonata, however, the cembalo discant actually is from the first
treble of the trio, the second treble being transferred to the gamba. This distribution is
not due to the gamba’s lower range, for in the trio the two trebles cross constantly and
occupy the same tessitura. Bach has deliberately placed the part of the first treble,
which has slightly greater dominance in the trio, into the cembalo. This distribution
gives the gamba the principal thematic material at the beginning of the first movement
(Ex. 7-1), but more importantly places the initial statements of the fugue themes in
movements two and four in the cembalo (Ex. 7-2). Had the parts been distributed

differently and the fugues begun with the theme in the gamba, the cembalo discant would

have been tacet and would have filled in harmony as a continuo.

 

5Rowen, Early Chamber Music, 166.

293

Ex. 7-1. 1039/10271, 1-3

1039

Flauto
traverse I

Flauto
tnn crso II

Camban

 

‘ _
======J --—-v I..- _.'

==__‘==-———_‘
_ —_— u===-_
-- _—

 

I. Adagio

   

Viola da gamba

Cembalo

A
rum-r 1r- m“ _.
l— — I‘l-‘.--.. " -
———-——=---.--— .4-5
-__ ———-—- _==-————--
--=

 

294
Ex. 7-2. 1039/1027 II, 1-8

1039
Allegro III non pesto

Ir

    

 

1027

2. Allegro ma non tanto

 

As the first flute part is moderately more brilliant than the second, particularly in the
second movement, the cembalo gains the more effulgent part in the transcription.

Further, since the gamba sounds an octave lower than the flute, the interweaving of

295

treble voices found in the trio is eliminated and the harpsichord discant gains uncontested
prominence as the top voice in the texture. In essence, the resulting dominance of the
cembalo reflects quite accurately Bach’s own title on the autograph: Sonata a Cembalo
e Viola da Gamba di J.S.B.‘5

Generally there are very few alterations made in 1027 to accomodate the changed
instrumentation. It is notable, however, that the tempo markings for the various
movements are altered in a manner comparable to that found in harpsichord concertos
1057 and 1062. In the latter works, transcribed from violin concertos, the slow middle
movements were modestly quickened while the tempos of some of the allegros were

moderated. The same pattern is evident in 1027, as illustrated by comparison with 1039:

Trio Sonata 1039 Gamba Sonata 1027
Adagio Adagio
Allegro ma non presto Allegro ma non tanto
Adagio e piano Andante
Presto Allegro moderato

The increased tempo for the third movement helps mitigate the harpsichord’s intrinsic
rapid tone decay, while the slowing of the finale assists the clarity of harpsichord tone
and articulation. Two additional accommodations may be perceived in the third
movement: the static repeated eighth notes in the trio’s continuo bass are replaced with
broken octaves which are more idiomatic to keyboard, an adaptation also found in the
D major concerto, 1054; and the short, two-note slurs of the second flute part are

replaced with smoother four-note slurs in the cembalo discant. These changes, including

 

”The title appears in Bach’s own hand on the autograph parts Mus. ms. Bach P 226.
This form of title is common in most of the sources for all of these sonatas. NBA KB
VH4, 11.

296

a comparison with the D major concerto, can be seen in Ex. 7-3. Noteworthy is the fact
that the articulation is altered only in the cembalo discant; the two-note slurs are retained
for the gamba.

Ex. 7-3, 1039111, 1-3

Adagio c piano

    
    

1027 III, 1-3

3. Andante

 

297
Ex. 7-3 (cont.)

1042 1, 4-5

 

Compared to the sometimes extensive alterations found in the concertos, there is
little of similar magnitude in this transcription. The changes or alterations of figuration
which do occur appear almost exclusively in the bass, the cembalo discant being virtually
unchanged from the Flute I part. Most commonly, changes in the bass involve brief
passages of sixteenth note diminutions which add a modest degree of brilliance and

rhythmic interest to the bass, as illustrated in the following examples (Ex. 74).

298
Ex. 7-4. 1039/1027, bass excerpts

1039 II, 14-16

 

.1
.

 

 

 

1027 II, 14-16

 

1039 II, 26—30

 

 

 

 

 

 

 

I I I
I I I I I I I I I I
x J 1 I I I I Jfi
4I I 1 I‘ I I I I—I
I- A I I A Ii]
7 A v
v

 

 

 

 

 

He

 

 

 

1027 II, 26—30

 

 

 

 

 

 

 

 

 

 

 

 

1039 II, 933-100
a; 7"?” ‘ o q

 

 

 

 

I
7: III III 71 I I

 

1027 11, 93-100

 

299

It is evident that Bach’s transcription process in this gamba sonata is very
straightforward, the upper flute parts of the trio being transferred virtually unchanged
to the gamba and cembalo while the continuo bass from the trio is adopted with minimal
changes. If both versions actually do go back to a still earlier original, as Eppstein has
proposed, then in all likelihood the reading of the original version was also the same,

since there is such considerable agreement between 1039 and 1027.

Parallel Versions Comparisons

In the absence of additional sonata pairs, i.e., sonatas with extant models, it
becomes particularly difficult to speculate whether the direct adaptation of 1039 to 1027
is typical of Bach’s transcription practice in the obbligato cembalo sonatas. However,
there is reason to believe that at least in principle the pattern of 1039/1027 was replicated
elsewhere. In the previous chapter, reference was made to a second version of the
middle movement of the A major flute sonata (1032) scored for violin, cello, and bass
and entitled ”Concerto." It is notable that the differences between this ”Concerto"
version and the flute sonata as found in the autograph are of the very same nature as the
modest changes found in 1039/1027: the upper voices are virtually unchanged while the
bass has a few modifications, generally diminutions as illustrated in Ex. 7-5. The
simpler reading of the ”Concerto" (source B) suggests that it is or reflects an earlier

version.7

 

7The ”Concerto” version of source B is reproduced in NBA KB VI/3, 55-57. The
disparity in measure numbers is due to the difference in meters: the flute sonata is in
6/8, the Concerto for violin, cello, and basso in 3/8. It should be remembered,
however, that the connection of this version with Bach is unknown.

300
Ex. 7—5. 1032 11, mm. 1-7, 1-14 sources A + B

A c dolce

 

 

The only other movement in the sonatas which allows a similar comparison
between parallel versions is the second movement of the G major violin sonata (1019).
As noted in chapter 7, there are three versions of this work, the earliest of which is

Mus. ms. Bach St 162 (source E in the NBA).8 For the first 10 bars of the second

 

8See NBA KB VI/ 1, pp. 139-140. Source E contains both violin and cembalo parts
for sonatas 1-4, but only the cembalo part for sonatas 5 and 6. St 162 is also a partial
autograph, Bach’s hand being evident in movements 3—5 of the sixth sonata.

301

movement, St 162 has a different reading of the cembalo bass: uniform quarter-note
motion with melodic intervals which outline triads reminiscent of continuo harmony (Ex.

7—6).

Ex. 7-6. 1019 II, 1-10 bass only (source E)

   
 

Largo

 

 

 

 

In the later version found in Altnickol’s copy (source A), the triadic outlines are replaced
with diminutions which provide a much smoother, scalar voice leading in eighth notes

(Ex. 7-7).

Ex. 7-7. 1019 II, 1-10 bass only (source A)

 

The basic harmonic outline remains the same, but the more expressive and interesting
voice leading of the later reading is not far removed from cembalo basses found in some
of the concertos, especially in solo episodes of concertos in which the cembalo bass is
freed from carrying the original bass (e.g., 1052 and 1058). It seems reasonable to
propose that if a sonata was derived from some other source, the bass in all likelihood

remained very close to the continuo bass of the original. This pattern is borne out in the

302

few examples cited above from sonatas with a model or alternate version. The same
principle is also visible in the concertos, except that in the latter the original bass may
be preserved intact in either the new continuo part or in the cembalo bass. When
entrusted to the continuo of the concerto, the cembalo bass is given greater freedom for
independent development. In a sonata, however, there is no separate continuo part, so
that any borrowed bass must be placed in the left hand part of the cembalo. As a result,
it is quite likely that the keyboard bass was more faithful to its source (if there was one)

in a sonata than in a concerto.

The Character of the Obbligato Parts

Frequently in the obbligato sonatas the left hand part suggests the character of
continuo bass. Traits typical of such bass lines include extended passages in uniform
eighth- or quarter-note rhythm which outline triads, or pedal points represented by static
repeated notes on a given pitch. Such passages are purely harmonic in function and
assume no melodic or motivic importance. It is noteworthy that such rudimentary bass
lines can and do appear in passages where the harpsichord functions as either continuo
or obbligato instrument, i.e., there is little or no stylistic differentiation made between
continuo and obbligato basses. As illustration, one need only look at the first movement
of the G minor gamba sonata where the same bass line appears first with figures for
continuo realization (mm. 3-9) and subsequently with a fully written out discant as an

obbligato part (mm. 13-19) (Ex. 7-8).

303

Ex. 7-8. 10291, 3-9

 

1029 I, 13-19

 

 

 

 

 

 

 

 

Passages of a similar nature are not difficult to find in the sonatas, and illustrate that bass
parts in the obbligato sonatas are often no different than the typical bass lines that had
long been the foundation of continuo practice. A parallel phenomenon may be observed
in harpsichord concertos where the continuo bass of a model is transferred to the

cembalo bass of the concerto (e.g., 1052a, 1054, and in all likelihood 1055 and 1056).

304

Inasmuch as the borrowed continuo bass is replicated as the bass for the solo cembalo,
the distinction between continuo and obbligato bass is again lost. In the concerto,
however, there is greater likelihood of at least modest embellishment, particularly in solo
episodes. On the other hand, in those concertos where the original continuo part is
transferred to the concerto continuo and the cembalo bass thus gains greater freedom
(e.g., 1052, 1058, and 1062), divergences from the continuo bass are distinguishable
since the latter is preserved in a separate continuo part. Most commonly, these
deviations are the result of embellishments or diminutions of the principal bass line. In
the sonatas, however, the absence of a separate continuo bass makes identification of
such alterations difficult if not impossible. Judging from diminutions observed in
1039/1027 and in other examples cited above, however, one might speculate that at least
some of the brief sixteenth-note figurations which decorate and fill in conjunct intervals
in passages of otherwise uniform rhythm may represent such embellishments. This
hypothesis does not presume that the underlying structural bass is drawn from some
previous source, but only that the character of the bass line is principally continuo-like
in nature, with occasional rhythmic/melodic embellishment to add interest and make the
part more idiomatic.

Mizler’s and Daube’s descriptions of Bach’s own thorough-bass accompaniment
cited earlier suggest that Bach’s discants drew upon thematic material from the solo part
so skillfully as to suggest a fully written out arrangement. Highlighted by both Daube
and Mizler was Bach’s reliance on melodic rather than chordal/harmonic material to
realize the discant, a description which characterizes fairly well the nature of the

cembalo discants in many of the obbligato sonatas. No longer limited only to continuo

305

harmony, the discant now participates fully and on equal terms with the solo. As noted
in the G major gamba sonata (1027), the discant was adopted virtually unchanged from
the second flute part of the trio (1039). Such direct transfer of a solo violin part to the
cembalo discant is a paradigm also found in harpsichord concertos such as 1054, 1058,
and 1062. In these works, the solo violin of the model was transferred directly to the
cembalo discant with only occasional added figuration or accomodation in keeping with
the keyboard’s technical capabilities. In the case of sonatas for which previous models
are unknown and may never have existed, perhaps the most that can be suggested is that
the cembalo discant often shares thematic and motivic material with the solo instrument
to such extent that the two become equal partners in the musical whole. It is for this
reason that Eppstein and others have so strongly proposed the possible or probable
existence of (lost) trio models for the obbligato sonatas, for in trios the two treble parts
usually exhibit this equality. The opening movement of the B minor flute sonata
provides an interesting example of this musical interplay. At mm. 32ff. the flute begins
a solo episode over a broken interval accompaniment, but at m. 37 the roles are reversed
as the cembalo discant restates the same solo material while the flute is given the broken
figuration previously heard in the cembalo, i.e. , stimmtausch (Ex. 79). This procedure

can be found replicated throughout this movement.

306

Ex. 7—9. 1030 I, 21-23, 27-29

21-23

27—29

     

   
  

‘ A —— --.‘ ».N
- u———. ——_--————
.--—--—-——u--'- ——-_—-oa—' -
Ila—lu_- n-l r- _._.. new I..-
\u_u-_'- I _1 r— n..—

This kind of musical discourse does represent a significant difference from that found
in a concerto, however, double concertos excepted. With two trebles, either or both
may assume the solo function, at least temporarily, whereas in a concerto the cembalo
discant retains its solo preeminence.

When considering the concertos earlier in this study, the use of chordal filler in

the cembalo in both tutti and solo passages was examined. It was suggested that the

307

occasional inclusion of chordal harmony in the cembalo represents lingering vestiges of
continuo practice, i.e., that the cembalo sometimes still retained its traditional function
of filling in the harmony. The obvious question which follows is whether such chordal
filler was employed at all in the obbligato sonatas. Daube’s description quoted earlier
noted Bach’s ability to imitate the upper voice even though "the regular accompaniment
was very little curtailed," suggesting that Bach’s own accompaniments included both
obbligato and continuo functions. In several of the obbligato sonatas, the occasional
inclusion of figured bass or chordal passages in the cembalo suggest that continuo
accompaniment is still employed. Examples can be found especially in 1030, 1016, and
1018. In four of the six violin sonatas (1014, 1015, 1018, 1019), figures appear under
the bass in several movements whenever the discant is tacet.9 The remaining two
sonatas (1016 and 1017) also present interesting examples. In 1017 there are no
measures in which the discant is silent, but the arpeggiated cembalo discant in the first

movement clearly replaces a chordal harmonization (Ex. 7-10).

Ex. 7-10. 1017], 1-3

Largo tr

 

 

9These figures are reprinted in the NBA in brackets, indicating that they come from
sources B-G. See NBA KB VI/l, 143.

308

The opposite circumstance, unfigured passages with tacet discant, appears in the finale
of 1016 at mm. 1-4, 35-37, 46-49, and 119-123. It is unthinkable that these passages
could be played without added continuo harmonization, despite the absence of figures

(Ex. 7-11).lo

Ex. 7-11. 1016 IV, 1—5

h n-"--- “-_"-— -- -—-----‘-
—-=—‘-——.-—-====_-.--—-‘.---——.-—-—.-—‘—
—-——-

m
.

 

 

W
n
T.

. .. s

More significant perhaps are passages where the harmony is fully realized as
written out chords within the cembalo parts. An obvious and most interesting example
appears in the first 12 bars of 1016 111. Here the cembalo discant plays nothing more

than repeated chords over an ostinato in the bass. In all respects, this is a fully written

 

“’I‘hat there was much discussion in eighteenth century writings about unfigured
basses illustrates that the absence of a figured bass was neither uncommon nor a
dismissal of the continuo. Heinichen complained of instrumental works "which
ordinarily have no figures over the bass, but one must oneself derive them extempore
from the score, indeed mostly from a single part above it, and must virtually guess the
rest by art and ear." Johann David Heinichen, Der General-Bass in der Composition
(Dresden, 1728), quoted in Rowen, Chamber Music, 51. Heinichen and C.P.E. Bach
both attempted to devise rules for such unfigured basses. See George J. Buelow,
Thorough-Bass Accompaniment according to Johann David Heinichen (Ann Arbor: UMI
Research Press, 1986), 219—220.

309

out continuo realization, above which the solo violin spins its florid melody in triplets
(Ex. 7-12). Most novel is the reversal of roles that commences at mm. 13 ff. as the .
cembalo takes up the violin’s florid melody while the violin accompanies with repeated
double notes in place of the cembalo’s previous chords. There is a certain paradox here

as the violin temporarily assumes the role of continuo for the cembalo.

Ex. 7—12. 1016 III, 1-16

_i—J —-:::-’ .1:-'v n-—-.‘.a m;—_.-
-1 n-- -—---——-u “Hum ":—- '—_- - — _ —
_-m- r' W
_—

-' -
--—.‘--' --' _-_. n--—---
I u——-—‘l -rv —: ..—-

r I
\.

 

A similar relationship between solo and discant appears in 1015 II where the cembalo

discant takes up a motive from the principal theme while the violin plays arpeggiated

310

chords in accompaniment. Particularly notable in this passage is the long pedal in the
bass on E, held for no less than 19 bars (and this in a movement which began
fugally).“ In essence, the violin and cembalo bass provide the harmonic framework

for the passage while the cembalo discant becomes the true solo (Ex. 7-13).

Ex. 7-13. 1015 II, 74—80

  

 

' O
=====E=EE=====EE=- ._..

 

Such reversal of roles between the two instruments is especially evident in 1018 II!
where the violin assumes the accompanying role not merely for a short passage but for
the entirety of the movement. Here the figuration is consistent from beginning to end,
producing a very unusual movement which is far removed from any traditional trio

sonata pattern (Ex. 7-14).

 

“Obviously the harpsichord’s tone cannot sustain for such a duration. As the tone
dissipates, the harpsichordist will need to restrike the octave. The customary trill on the
note is unsatisfactory if both pitches of the octave are to sound.

311
Ex. 7-14. 1018 III, 1-5 (BG)

Adagio.

 

 

 

 

. _ ' -_.
m-—W . ---. r7—
.-‘— ,, run-iv —— A44 n-v ru—n'uu— local

a" --_ u c'—— === . , —==="?T' '._"'_ "4.. "'",'§§-_.__—

 

A variant reading of the cembalo part for this movement is contained in several of the
earlier sources for the violin sonatas.12 In this alternate version (Ex. 6-19), the
cembalo parts appear as arpeggiated triads in sixteenthmote motion. Given the static
chordal repetition of the violin part, the embellished later reading with figuration in
32nds shown above is considerably superior and encompasses a recurrent motive in the
discant which helps unify the movement considerably. The filling in of arpeggiated

triads also further illustrates Bach’s technique of diminution.

 

l2Sources C, D, and E in the NBA, identified as "Weyses samling," (Kgl. Bibliothek
Kopenhagen), Am.B. 61, and Mus. ms. Bach St 162 respectively. See NBA KB VI/ 1,
138-140 and 198. This alternate reading is reproduced in the appendix to NBA VI/ 1,
195-196.

312

The only obbligato sonata outside the violin sonatas to use chordal filler is the B
minor flute sonata (1030). In three of its four movements appear passages with chordal ‘
harmony in the discant, and in each instance the cembalo functions as continuo
instrument while the flute carries the theme. The finale to this sonata is most unusual
in that it is in two sections, the first a fugue, the second a gigue in 12/16 meter. The
syncopated theme which opens the gigue is played over a written out continuo
accompaniment (Ex. 7-15). Except for a very brief return of this chordal
accompaniment in the second half of the binary form, the remainder of the gigue finds
cembalo and flute sharing the musical material equally. Thus both types of cembalo

playing--obbligato and continuo—-are juxtaposed again in the same movement.

Ex. 7-15. 1030 III, 84-87

84

    

 

313

Special Features of the Violin Sonatas

As a group, the violin sonatas exhibit certain features which distinguish them
from the other sonatas with obbligato cembalo. Several of the foregoing examples from
the violin sonatas have already highlighted a new interdependent relationship between
solo and cembalo. Two styles of composition, the learned and free, may be identified
in these sonatas. The first is derived from the trio sonata, with three active parts
represented most clearly in the fugal movements which comprise all the allegros of the
first five sonatas (1014-18) plus the finale of 1019. In each of these movements, the
theme of the fugue finds its way eventually into the bass, so that the bass obtains the
double function of presenting thematic material and also of providing the bass of the
harmony. The second type, free composition, is represented in most of the slow
movements and is more progressive.13 In these movements can be found exemplars of
the true accompanied solo, in which discant and violin are thematically independent.
The boldest step into this manner of writing is found in the first movements of 1014 and
1018. In the latter, the cembalo features three-part imitative polyphony built around a
recurrent motive that never appears in the solo violin. The expansive length of this
movement, 106 bars in Largo, places it beyond all other slow movements in the chamber
sonatas. The first movement of 1014 finds an expressive solo line in the violin
accompanied by double thirds and sixths in the discant over an ostinato-like figure in the

bass. Each part is self-contained, and invites no further harmonic elaboration. In the

 

]”The slow movements in 1015 and 1019 are imitative, adhering to trio sonata style,
though without the appearance of the theme in the bass. The top two voices in 1015 III
are even in strict canon throughout, a clear indication of the style from which this work
derives.

314

middle of this movement, the double thirds/sixths become the principal musical material
for a shared dialogue between violin and discant. This section is particularly notable for
the interplay between the two treble voices. While the musical material is clearly
related, it is not imitative; what one part begins, the other continues, weaving the line
into a beautifully unbroken and expressive cantilena until a common cadence is reached
at m. 20, at which point the process begins anew (Ex. 7—16). The traditional solo-
accompaniment model thus is replaced by two equal partners in a duet texture

comparable only to the solos of a double concerto or duet aria.

Ex. 7—16. 10141, 16-21

16

 

Comparisons between Sonatas and Concertos
That some movements in these violin sonatas should invite comparison with the
harpsichord concertos is not surprising, particularly in the case of the accompanied solo

type movements. Sonata movements in which the violin is accompanied by the cembalo

315

suggest some influence from the earlier solo sonata (with continuo) and from the violin
concerto, while the reverse distribution of cembalo as solo accompanied by violin implies
procedures drawn from the solo harpsichord literature as well as the harpsichord
concerto. The most obvious outward parallel between a sonata and concerto perhaps is
represented by 1017 I and'1053 11, both Siciliano movements. The opening measures

especially show distinct parallels (Ex. 7-17).

Ex. 7—17. 1017 I, 1-3 and 1053 II, 1-2

 

_-
—--———— --—--—--——— -—--
—-- _—---- -==———- _———-r—-—
-—- — -=:—'-.- —:
--_: — r-—-

 

1053

 

There are striking similarities present here, including the singing cantilena in the solo
violins, both with the same dotted rhythm patterns, over a very keyboard-idiomatic

arpeggiated accompaniment. Both examples are also in minor keys, and illustrate the

316

"accompanied violin” type slow movement. The relationship between parts changes in
the concerto at m. 7, however, when the cembalo discant asumes the solo role against
a light chordal accompaniment in strings (Ex. 7—18).

Ex. 7—18. 1053 II, 6-11

I a
.- ——-——_——_
a

—— ———_

r— _-‘-_—

-—-—-.—. 1 l-“—-- . L——-—— v—_—u—————_——-—_
_ _—

. —— _—
’La _ — ~—. u— _.‘. -_ _ . -1.—

. .--———.—-—————..-—_""

- _——— '—

 

This passage and what follows has no counterpart in 1017, but it does parallel somewhat
the pattern of the cembalo accompanied by violin found occasionally in the sonatas. The
remainder of the concerto movement is considerably more developed than the Siciliano
in the sonata, exploring richly varied melodic embellishments consonant with its concerto

character.

317

The second movement of the D major harpsichord concerto (1054) contains a
somewhat similar example in which the solo material appears in the cembalo discant
while repetitive block chords in the strings and cembalo bass accompany in continuo
fashion (Ex. 7-19). Parallels with passages in 1018 III (Ex. 7—14) and 1016 III, 13-20

(Ex. 7-12) are striking.

Ex. 7-19. 1054 II, 28-31

 

In the E major violin concerto (1042) which served as model for 1054 there is no
continuo bass in the passage just cited, and so the harpsichord bass doubles the viola in
1054. Thus strings and cembalo bass become the 'continuo" for the harpsichord discant,
much as in the passage cited from the third movement of 1016.

There are several additional parallels that may be drawn between sonatas and
concertos, but considerable space need not be devoted to them here. First, the
appearance of figured basses in sonatas when the discant is tacet is a feature also found
in the concertos, and under the same conditions, i.e., when the right hand part of the
cembalo is vacant. The number of such passages with figured basses is far less in the

concertos than in the sonatas since the cembalo discant is kept active virtually all of the

318

time in concertos, either doubling violins in ritornellos or playing solo episode
material.“ Second, as noted in chapter III, several of the concertos were modeled after
violin concertos, the cembalo discant being derived directly from the solo violin of the
model. The style of writing in the cembalo discants in the violin sonatas often suggests
a similar transference of figuration idiomatic to violin, which is why Eppstein’s
hypothesis of trio sonata origins is not altogether unreasonable. As noted in 1039/1027,
the first violin part of the trio sonata became the discant of the cembalo, the same
transfer pattern found in the concertos arranged from violin concertos (1054, 1058, and
1062). Third, the most florid and idiomatic writing for violin in the violin sonatas, such
as 10161, finds a parallel in the equally florid keyboard writing in the second movement
of the F Minor Concerto (1056); both movements well illustrate Bach’s mastery of

diminution, and hint at its archetype (Ex. 7-20).

Ex. 7-20. 10161, 3-6, and 1056 II, 9-12

1016

 

 

1‘Examples include 1054 II, 1-5; 1055 II, 1-2, 7-8, 37—38. The recurrence in 1054
II, 51-57, of the same bass as mm. 1-5 would suggest that the same continuo realization
should be followed even though no‘figures appear.

319
1056

a

wn-———— _——— _—

- ~_-—4——_— ' ‘—'———--.——-'v— _—4

:5 _ . -—-—— —
——_— _ _— "

-‘ _4 an-

"_.‘...F

-- c ‘ ——.-— ———='==
‘—

 

Finally, it will be remembered that several of the sonatas, including 1032, 1029,
and 1019 (for flute, gamba, and violin, respectively) suggest the concerto in their very
design, at least in certain movements. This resemblance is reflected in the number of
movements and their tempo relationships as well as in the implied use of ritornellos. It
was noted also in chapter 7 that the source materials for several if not most of the
sonatas, at least in their final versions, are traceable to Leipzig, some datable to the
period of the Collegium and hence within close temporal proximity to the concertos.
This recognition inevitably leads to the question whether the concertos, especially the
multi—keyboard concertos which originated first, might not have influenced the sonatas.
No definitive answer may be possible, but given the influx of concerto elements into the
sonatas, the prospect of such cross influence is very inviting and altogether likely.
Consider that the A major flute sonata (1032), which particularly suggests concerto
design in its opening movement, is securely datable to 1736 and is preserved alongside
the C minor double harpsichord concerto in the same autograph. The date places the

work in the middle of Bach’s first tenure with the collegium, and since the concerto

320

obviously was written into the manuscript first it can be verified that the sonata was
c0pied afterwards by Bach into the same manuscript; i.e., the concerto preceded the
recopied/reworked sonata. It is not hard to envisage that Bach’s renewed interest in and
experience with concerto composition, put to practice in the double harpsichord
concertos, may have also found fertile ground in contemporaneous sonata composition.

Before leaving the sonatas, a few final comments concerning the G major violin
sonata (1019) are in order. As noted in chapter 7, this sonata exists in three versions,
the earliest of which is datable with some assurance to 1724/1725. The final version,
preserved in a copy by Altnickol, comes from no earlier than 1744. The first two
movements are common to all three versions, but the last three movements are new in
the final version, and hence clearly were written in Leipzig. On stylistic grounds,
Eppstein has suggested that this sonata is the earliest of the six violin sonatas, despite its
last position in every manuscript copy.ls Eppstein’s conclusion was driven by his
judgement that the four-movement sonata type incorporating tutti-fugues and
accompanied solos represented the summa of Bach’s sonata output; since the violin
sonatas as a group had to be placed last among all the sonatas for similar reasons, the
sixth sonata thus must come before the remaining five, all of which are in four
movements with tutti-fugues and the like. However, since the first five sonatas are
virtually the same in all manuscript sources (i.e. , they are unchanged between source ’E’
of 1724/25 and source ’A’ of the 1740’s), the G major sonata, being the only work with

extensive revisions and reworkings, actually becomes the latest violin sonata, at least in

 

l5Eppstein, ”Studien,” 156. The ordering of the sonatas--B minor, A major, E
major, C minor, F minor, and G major--is consistent in all sources.

321

its final form. What is particularly striking about the violin sonatas and especially this
G major sonata is the brilliance of the writing for the cembalo. As already noted, the
cembalo bass often strongly suggests derivation from a continuo bass, with occasional
embellishment, diminutions, and accomodation to the technique of the keyboard. In both
the opening and closing movements of 1019 the left hand part is notably brilliant, with
far less overt suggestion of continuo derivation. In the finale, which is definitely a
product of Leipzig, the bass is replete with running passages which imitate or sometimes
even introduce motives that subsequently dominate the musical material of all parts.
Although this movement .is a tutti-fugue, the inclusion of such motivic material in the
bass is not necessitated by fugal principles, since the sixteenth-note figurations are not
part of the fugue theme. The introduction of such a motive, first in the bass and

subsequently in upper parts, may be illustrated from mm. 14 ff. (Ex. 7-21).

Ex. 7-21. 1019 V, 14-20

 

322

The opening movement, though originating in the first version of the sonata (1724), is
equally brilliant in its fashioning of cembalo parts. As noted earlier in Chapter 7, the
structural design of this movement is very unusual, combining elements of fugue and
concerto. Of greater interest here, however, are the brilliant passages for left hand
which are very idiomatic to the keyboard and are removed from typical continuo-like

patterns (Ex. 7-22).

Ex. 7-22. 10191, 16-21, 63-68

16f.

-—-——————;—1 .- _
- -—-—_-——-—r' ..._._-_-:=
_- _._.—

" J" — n m .——
- "A—un-‘--.‘-—

---u--—--—.-I--——-——-
-_--———_——--4-——__#—
‘ ._._ _fi ”-

63f.

..— - -
m— _ -_-'—--
-l-'_:=Hqu-"r=

— m-—-—‘— — ---m

a '__.‘d r-l - r—-—-—=————r_“‘-

— “i __

---.r-_" .— — r. r-‘ u
=' A'_ ‘fl—-‘ "----' I“-‘ u--=-
-'—.". =4:==.'==='—."" ‘M-'-"7-‘
:_ 'r— u-_——-—-—--'_—-

 

  
   
  

   

—‘ 4
—_-———5 '

= . =-—_.:
--— m——':--

 
 
       

 

 

l
-v- -1 r- -‘-—— —_'—I
'.—_—-‘—-————_~—-_
u.=_————_.———-————

    

—. .--—--—_-_—- —_=::_
-----—:_,——__—-—-.-- '_——l \
I-- -"_‘—-__-J '-----—

_‘ —-——‘-.—_-
—_~» ——_. '---_--_-‘-'_

  
 
 

  
  
 

.—. =;-—~-
_.‘ -_- ---
E_._.—”fl." ..
\:__:_... .-

323

Particularly in the last excerpt cited, the exploitation of keyboard idiom is evident in both
bass and discant; the discant even includes a brilliant descending scalar passage extending
across two and a half octaves.

Although the first and last movements of this sonata are separated by perhaps as
much as twenty years, it is noteworthy that there is no disparity nor lessening of intensity
so far as the writing for cembalo is concerned. In its final form, 1019 is perhaps the
most singular of all the sonatas, being comprised of five movements, the central one of
which dispenses with the solo instrument altogether and instead presents a piece for
cembalo alone in the manner of a suite movement but preserving the three part texture
of the other movements. Needless to say, the writing is highly idiomatic to the
keyboard. Collectively, 1019 in its final form brings together in summary fashion an
astounding variety of musical types: implied ”Concerto allegro,” tutti-fugue, solo suite,
three-part fugal, and all with a mix of continuo- and obbligato-type cembalo writing.

Given the lack of extant models or parallel versions for the sonatas, it becomes
much more difficult to draw definitive conclusions about Bach’s creative and
compositional process in these works. Nonetheless, similar goals of elevating the
cembalo to a position of ascendancy while simultaneously moving it away from its
continuo role are evident in both the sonatas and concertos. The left hand part, so
crucial to this transformation, is perhaps more closely tied to its continuo background
in sonatas than is true in the concertos, in part owing to the absence of a separate
continuo bass part. Visible traces of embellishment in the bass, however, suggest similar

procedures in fashioning the cembalo bass. In both sonata and concerto, the continuo

324

bass clearly served as progenitor of the cembalo bass, forming an important point of

departure for the development of new obbligato parts.

Summary and Conclusions

The foregoing chapters have examined a significant number of Bach’s
instrumental ensemble works, including 15 concertos, 11 sonatas, and 4 cantatas, plus
additional alternate readings or secondary versions as appropriate: more than 30
compositions in all. The central purpose has been to study these works individually and
collectively with particular attention to Bach’s transformation of the harpsichord from
continuo into a solo instrument, a change of function which was foundational to the
shaping and emergence of the keyboard concerto.

All of these harpsichord concertos are transcriptions of earlier works, and in each
instance the original version apparently was a concerto as well; that is, these concertos
represent a transcription between like works, concerto to concerto, albeit with changed
instrumentation. Related cantata movements, then, are recognized as parallel versions,
not originating models, inasmuch as they represent reworkings of earlier, now lost,
concerto originals. This judgment rests largely on the formal designs of these cantata
movements, which bear the imprint of features commonly associated with the concerto.

The harpsichord concertos are all products of Leipzig, and stem from Bach’s
association with the collegium musicum that he directed from 1729-37 and again in 1739-
41. Contemporary records of the collegium and its performance schedule affirm that
sufficient performers, many drawn from the ranks of University students, were available

to present these works, particularly when abetted by ”traveling virtuosos" who

325

326

frequented the Bach household during visits in Leipzig. The concertos for multiple
keyboards were composed first, and in all likelihood involved Bach’s sons or students
in their performance. While the sonatas with obbligato cembalo are dated less securely,
extant source materials place all of them, at least in their final form, also in Leipzig; the
flute and gamba sonatas appear to stand in close temporal proximity to the concertos.
In this light it is notable that the B minor flute sonata and especially the gamba sonatas
require the same keyboard gamut--GG to d”’--as appears in several of the single
harpsichord concertos (1053, 1054, 1055, 1056).

It is likely that both the concertos and at least some of the sonatas shared a
common means of provenance: both were derived via transcription from earlier (extant
or lost) models. Given the considerable lack of surviving source materials for original
versions of the sonatas, however, determination of whether a sonata is derived by
transcription or not is very difficult if not impossible to ascertain. Sonatas which betray
concerto-like characteristics (e. g., 1029, 1032) are by their very design most attractive
candidates for such conjectures, but also possible is the hypothesis that such sonatas may
represent examples of the Sonate auf concertenart described by Scheibe. More
importantly, however, it cannot be presumed that all of these sonatas necessarily arose
as the result of transcription from earlier readings. This common assumption has been
accepted widely by many scholars despite lacking documentation; such a premise may
lead to faulty conclusions especially about dating and chronology.

The common denominator between these concertos and sonatas is, of course, the
obbligato cembalo. In both concerto and sonata, the employment of a fully realized

harpsichord implies a growing recognition of the weakening importance of basso

327

continuo, the immutable prerequisite of the age. It is perhaps not surprising that sonatas
for cembalo with melody instrument are among the first works to relinquish the
omnipresent continuo, since the keyboard instrument, even when cast in a solo role,
remains uniquely capable of simultaneously realizing the harmonic function which was
the foundation of continuo practice. In the concerto, the retention of the borrowed
continuo bass in either the cembalo left hand part or in a separate continuo part affirms
that the continuo was not really abandoned, but also that the role of the harpsichord was
considerably changed. That figures continued to appear under the bass in passages
where the cembalo discant is tacet indicates that the harpsichordist sometimes fulfilled
a double responsibility of soloist and thorough-bass player.

Despite such lingering vestiges of continuo practice, the solo cembalo nonetheless
is fashioned into a new role, first via a fully written out discant derived from the
principal string part(s) of the model, and second by elaboration and embellishment of the
bass. Greater attention is given to the latter in concertos than in sonatas, though the
degree of embellishment varies widely. From the autograph of the concertos it is evident
that Bach worked out the cembalo bass and continuo bass simultaneously, thus giving
careful consideration to the function and identity of both bass parts. Particularly in the
sonatas, continuo bass and cembalo bass are in essence one and the same, the continuo-
like character of the bass preserved with relatively few alterations. A somewhat parallel
circumstance is found in those concertos where the borrowed bass from a model is
placed principally in the left hand part of the cembalo. In such instances, the cembalo
bass and original bass coincide. On the other hand, the development of a largely

independent left hand part is requisite to the full establishment of this new genre, and

328

steps in this direction are apparent more in the concerto than in the sonata. In at least
some concertos (1052, 1058, 1060, and 1062), the cembalo bass demonstrates greater
independence and freedom from the continuo bass, and these same works also exhibit
the greatest degree of brilliance and idiomatic writing for keyboard. It must be noted,
however, that concertos of this latter type were written alongside concertos (and sonatas)
in which the cembalo bass duplicates or is closely tied to the continuo, i.e. , no hierarchy
should be imagined, for both types existed side by side.

Somewhat similarly, two types of sonatas may be distinguished. The first is
closest to the traditional trio sonata, with four movements incorporating fugal allegros,
while the second type is freer, often comprised of three movements in which the allegros
sometimes bear the imprint of concerto form. As with the concertos, both types existed
side by side and without prejudice. Some of the most forward-looking writing in the
sonatas is found in certain slow movements which are cast as true accompanied solos.
In these a new level of independence between solo instrument and cembalo is attained,
marked by complete autonomy of thematic material.

Finally, although the degree of cross influence between obbligato sonata and
concerto is difficult to ascertain, there are a number of observable common features
which suggest that one genre may well have informed the other. Certainly the kinds of
alterations Bach made when transcribing the concertos are also found in the sonatas:
accomodation of string idioms, alterations for range, exchange of parts, arpeggiation and
ornamentation (diminutions) to add brilliance, octave displacements (unrelated to range
problems), and added chordal filler. These demonstrate a shared transcription

methodology, a common basis upon which works for similar instrumentation-~strings (or

329

flute) and cembaloucould be fashioned. While the keyboard writing is much less
brilliant in the sonatas than the concertos, it nonetheless remains that the sonatas still
may have provided a proving ground for works in which the cembalo no longer functions
as continuo. Fashioning a cembalo discant so as to suggest ritornello and solo episodes,
as in 1019 and 1029, finds direct parallels with the cembalo writing in the concertos.
Perhaps more intriguing is the notion that the concerto may have influenced the sonata,
first in its formal design and emphasis on the solo part, but also in slow movements
where the cantilena style exploited in concerto slow movements is emulated. The
temporal proximity of all these works, plus their likely side-by-side performance in the
collegium, lends support to such conjectures about mutual influence. That some scholars
have found certain of these sonatas to exemplify the Sonate auf Concertenart only
highlights the extent of mixed sonata and concerto elements in these works. At the same
time, it is perhaps too much to assert that Bach learned from the sonatas how to write
for cembalo in a concerto setting; but there are certainly enough commonalities to
strongly suggest mutual cross influence.

There remain many unanswerable questions, in large measure due to the
regrettable lack of autograph materials. First and foremost, further knowledge of
original models, which would verify the transcription status of the sonatas especially, is
unlikely unless additional sources come to light. Much of the present literature about
these works is dominated by discussions of the presumed original models and their
conjectured instrumentation and keys. Related questions concerning octave and key
transpositions also remain clouded. Second, there are questions concerning dating and

chronology of the obbligato sonatas, issues which relate to cross influences with the

330

concertos. Third, there is the question of a second harpsichord to play continuo. There
is nothing that supports the use of continuo in the sonatas, but its employment in the
concertos remains debated. Fourth, the issue of a violone as a doubling bass instrument
is not documentarily proven, despite references to ”violone” in 1057 and a separate part
for 1055. If such an instrument were admitted to the instrumentarium, there remains the
question of which violone is intended. Such questions, and many more, find no
definitive answers, but have opened the way to considerable conjecture and speculation.
If nothing else, they have summoned considerable discourse and hypothesis in the current
Bach literature.

What remains undisputed is that these works were brought to life by Bach at an
expedient point in time when changes of far-reaching import were taking place. That
the harpsichord should arise in new guise and usher in a new genre at the very time
when it itself was about to be eclipsed by the fortepiano, the instrument which would

inherit and fully realize this development, must remain one of music’s great ironies.

BIBLIOGRAPHY

Bibliography

Music Editions

Johann Sebastian Bach, Neue Ausgabe sa'mtlicher Werke (NBA). Edited by the Johann-
Sebastian-Bach-Institut, Gottingen, and the Bach Archiv, Leipzig. Kassel: Barenreiter;
Leipzig: Deutscher Verlag fiir Musik, 1954-.

I/ 11.2 Kantaten zum Sonntag Jubilate. Edited by Reinmar Emans. 1989.

V/I

V/4

VI/ 1

VI/3

VI/4

VII/2

VII/3

VII/5

Erster Teil der Klavierfibung. Edited by Richard Douglas Jones. 1976.

Das Klavierbiichlein fiir Anna Magdalena Bach. Edited by Georg von
Dadelsen. 1957.

Werke ft'ir Violine. Edited by Rudolf Gerber. 1958.
Werke ffir Flote. Edited by Hans-Peter Schmitz. 1963.

Drei Sonaten fiir Viola da Gamba und Cembalo. Edited by Hans Eppstein.
1984.

Sechs Brandenburgische Konzerte. Edited by Heinrich Besseler. 1956.

Konzerte fiir Violine, fiir zwei Violinen, fiir Cembalo, Flote und Violine.
Edited by Dietrich Killian. 1986.

Konzerte fiir zwei Cembali. Edited by Karl Heller and Hans-Joachim
Schulze. 1985.

VII/7 Verschollene Solokonzerte in Rekonstructionen. Edited by Wilfried Fischer.

1970.

Johann Sebastian Bachs Werke (BG). Edited by the Bach-Gesellschaft zu Leipzig. 47
vols. Leipzig: 1851-1899; reprint Ann Arbor: J.W. Edwards, 1947.

9

Kammermusik: Drei Sonaten fiir Clavier und Flote. Sechs Sonaten fiir
Clavier und Violine. Suite fiir Clavier und Violine. Drei Sonaten fiir

331

 

332

Clavier und Viola da gamba. Sonate fiir Flote, Violine und bezifferten Bass.
Sonate fiir zwei Violinen und bezifferten Bass. Edited by Wilhelm Rust.
1860.

10 Kirchenkantaten 41-50. Edited by Wilhelm Rust. 1860.

17 Sieben Concerte fur Clavier mit Orchesterbegleitung. Edited by Wilhelm
Rust. 1869.

32 Kirchenkantaten 151-160. Edited by Ernst Naumann. 1886.
33 Kirchenkantaten 161—170. Edited by Franz Wfillner. 1887.
37 Kirchenkantaten 181-190. Edited by Alfred Dorffel. 1891.

Sonata in F major for Clavier and Violin, BWV 1022. Edited by Ludwig Landshoff.
Leipzig: C.F. Peters, 1936.

The Six Brandenburg Concertos and the Four Orchestral Suites. From the Bach-
Gesellschaft Edition. New York: Dover Publications, 1976.

Complete Concerti for Solo Keyboard and Orchestra. From the Bach-Gesellschaft
Edition. New York: Dover Publications, 1985.

The Three Violin Concerti. From the Bach-Gesellschaft Edition. New York: Dover
Publications, 1986.

Literature

Aber, Adolf. "Studien zu J. S. Bachs Klavierkonzerten." Bach Jahrbuch 10 (1913): 5-
30.

Adlung, Jacob. Musica mechanica Organoedi. Berlin, 1768. Facsimile edition edited
by Christhard Mahrenholz. Documenta Musicologica Erste Reihe: Druclcschnfien-
Faksimiles 18. Kassel: Barenreiter, 1961.

Aldrich, Putnam. ”Bach’s Technique of Transcription and Improvised Ornamentation. "
The Musical Quarterly 35 (1949): 26—35.

Bach, Carl Philipp Emanuel. Versuch iiber die wahre Art das Clavier zu spielen.
Facsimile edition edited by Lothar Hoffmann-Erbrecht. Leipzig: Breitkopf und Hartel,
1969.

333

Besseler, Heinrich. Kritischer Bericht to NBA VII/2, Sechs Brandenburgische Konzerte.
1956.

------- . "Zur Chronologie der Konzerte Joh. Seb. Bachs. " Festschnfi Max Schneider.
Edited by Walther Vetter, 115—128. Leipzig: Deutscher Verlag fiir Musik, 1955.

Breig, Werner. "Johann Sebastian Bach und die Entstehung des Klavierkonzerts.”
Archiv fiir Musikwissenschaft 36, no. 1 (1979): 21-48.

------- . ”Probleme der Analyse in Bachs Instrumentalkonzerten. " Bachfest—Symposium
I978 der Philipps- Universita't Marburg. Kassel: Barenreiter, 1981.

------- . ”Zum Kompositionsprozess in Bachs Cembalokonzerten. " J. S. Bachs Spatwerk
und dessen Urnfeld: Prospektiven und Probleme. Edited by Christoph Wolff, 32-47.
Kassel: Barenreiter, 1988.

------- . "Zur Chronologie von Johann Sebastian Bachs Konzertschaffen." Archiv fit‘r
Musikwissenschaft 40 no. 2 (1983): 77-101.

Buck, Charles H. III. ”Revisions in Early Clavier Concertos of C.P.E. Bach:
Revelations from a New Source. " Journal of the American Musicological Society
(1976): 127-132.

Buelow, George J. Thorough-Bass Accompaniment according to Johann David
Heinichen. Ann Arbor: UMI Research Press, 1986.

Burney, Charles. A General History of Music from the Earliest Ages to the Present
Period. Vol. 2. 1789. Reprint, with critical and historical notes by Frank Mercer,
New York: Dover Publications, 1957.

Dadelsen, Georg von. Beitra'ge zur Chronologie der Werke Johann Sebastian Bachs.
Tiibinger Bach-Studien 4/5. Trossingen: Hohner, 1958.

David, Hans, and Arthur Mendel. The Bach Reader. Revised edition. New York: W.
W. Norton and Company, 1966.

Dreyfus, Laurence. ”The Articulation of Genre in Bach’s Instrumental Music.” The
Universal Bach, 10-37. The American Philosophical Society, 1986.

------- . Bach ’s Continua Group. Cambridge: Harvard University Press, 1987.

------- . "J. S. Bach and the Status of Genre." The Journal of Musicology 5 ( 1987): 55-
78.

334

------- "J. S. Bach’s Concerto Ritornellos and the Question of Invention." The Musical
Quarterly 71 (1985): 327- 58.

------- . ”The Metaphorical Soloist. Concerted organ parts in Bach’s cantatas." Early
Music 13 (1985): 237-247.

Dreyfus, Laurence, ed. Concluding Remarks in 3 Sonaten firr Viola da Gamba und
Cembalo. Leipzig: C. F. Peters, 1985.

Drummond, Pippa. The German Concerto: Five Eighteenth- Century Studies. Oxford:
Oxford University Press, 1980.

Diirr, Alfred. Kritischer Bericht supplement to NBA VI/3, Sonata A-durfiir Flauto
Traverso und Cembalo BWV 1032. 1981.

------- . ”Zur Chronologie der Handschrift Johann Christoph Altnikols und Johann
Friedrich Agricolas." Bach Jahrbuch 56 (1970): 44-65.

------- . ”Zur Chronologie der Leipziger Vokalwerke J. S. Bachs." Bach Jahrbuch 44
(1957): 5-162.

Eller, Rudolf. "Uber Charakter und Geschichte der Dresdner Vivaldi-Manuskripte. "
Vivaldiana 1 (1969): 57-63.

------- . ”Vivaldi-Dresden-Bach. " Beitrage zur Musikwissenschaft 3 (1961): 31-48.

------- . "Zur Frage Bach-Vivaldi. " Bericht fiber dem Internationalen
Musikwissenschaftlichen Kongress Hamburg 1956. Edited by Walter Gerstenberg, 80-
85 . Kassel: Barenreiter, 1957.

Eller, Rudolf, and Karl Heller. Kritischer Bericht to NBA VII/6, Konzertefiir drei und
vier Cembali. 1976.

Emans, Reinmar. Kritischer Bericht to NBA 1/ 11.2, Kantaten zum Sonntag Jubilate.
1989.

Emery, Walter. "A Neglected Bach Manuscript." The Music Review 11 (1950):169-
174.

Eppelsheim, Jfirgen. "Beobachtungen am Instrumentarium und Orchester Bachscher
Kompositionen aus den beiden letzten Lebensjahrzehnten. " J. S. Bachs Spatwerk und
dessen Urnfeld: Prospektiven und Probleme. Edited by Christoph Wolff, 77-83.
Kassel: Barenreiter, 1988.

335

Eppstein, Hans. Kritischer Bericht to NBA VI/4, Drei Sonatenfiir Viola da Gamba und
Cembalo. 1989.

------- . ”Grundzfige in J. S. Bachs Sonatenschaffen." Bach Jahrbuch 56 (1969): 5-30.

------- . "J. S. Bachs Triosonate G-dur (BWV 1039) und ihre Beziehungen zur Sonate fiir
Gambe und Cembalo G-dur (BWV 1027).” Musilq‘orschung (1965): 126—137.

Eppstein, Hans. "Studien iiber J. S. Bachs Sonaten fiir ein Melodieinstrument und
obligates Cembalo. " Acta Universitatis Upsaliensis: Studio musicologica Upsaliensia
2. Uppsala: Almquist & Wiksells, 1966.

------- . "Uber die Beziehungen zwischen Konzert und Sonate bei Johann Sebastian
Bach. " Beitrage zum Konzertschafien Johann Sebastian Bachs. Bach-Studien 6.
Edited by Peter Ahnsehl, Karl Heller, und Hans-Joachim Schulze, 87-94. Leipzig:
Breitkopf und Hartel, 1981.

------- . ”Zur Problematik von J.S. Bachs Sonate filr Violine und Cembalo G-dur (BWV
1019)." Archiv fiir Musikwissenschaft (1964): 217-242.

Fischer, Wilfried. Kritischer Bericht to NBA VII/7, Verschollene Solokonzerte in
Rekonstruktionen. 1971.

Fischer, Wilhelm. "Zur Entwicklung der Wiener klassischen Stils.” Studien zur
Musikwissenschaft 3 (1915): 24-84.

Forkel, Johann. Johann Sebastian Bach: his life, art, and work. 1802. Reprint, New
York: Da Capo Press, 1970.

Geiringer, Karl. Die Musikerfamilie Bach. Munich: C. H. Beck, 1958.

Glockner, Andreas. "Neuerkenntnisse zu Johann Sebastian Bachs Auffiihrungskalender
zwischen 1729 und 1735.” Bach Jahrbuch 67 (1981): 66-75.

Hausswald, Gunter. "Bach und der Dresdener Komponistenkreis." J. S. Bach, 18-23.
Dresden: Dresdener Verlagsgesellschaft, n.d.

Hauswald, Giinter, and Rudolf Gerber. Kritischer Bericht to NBA VI/l, Werke fiir
Violine. 1958.

Haynes, Bruce. Music for Oboe, 1650-1800. A Bibliography. Berkeley: Fallen Leaf
Press, 1992.

Heller, Karl, and Hans-Joachim Schulze. Kritischer Bericht to NBA VII/5, Konzertefiir
zwei Cembali . 1990.

336

Hering, Hans. ”Bachs Klavieriibertragungen. Ein Beitrag zur Klavieristik.” Bach
Jahrbuch 45 (1958): 94-113.

Herz, Gerhard. ”Yoshitake Kobayashi’s Article ’On the Chronology of the Last Phase
of Bach’s Work: Compositions and Performances, 1736 to 1750:’ An Analysis with
Translated Portions of the Original Text.” Bach 21 (1990): 3-25.

Hiller, Johann Adam. Lebensbeschreibungen beru‘hmter Musikgelehrten und Tonkunstler
neurer Zeit. Leipzig: Edition Peters, 1975.

Hirsch, Paul. ”Uber die Vorlage zum Klavierkonzert in d-moll." Bach Jahrbuch 26
(1929): 153-174.

Hoffmann-Erbrecht, Lothar. ”Die Bedeutung der Konzertform Bachs fiir die Geschichte
des Solokonzerts. " Bericht iiber die Wissenschaftliche Konferenz zum III.
Intemationalen Bach-Fest der DDR, Leipzig, 18./I9. September 1975. Edited by
Werner Felix, W. Hoffmann, and A. Schneiderheinze, 281-286. Leipzig: VEB
Deutschen Verlag fiir Musik, 1977.

------- . ”Johann Sebastian Bach als Schopfer des Klavierkonzerts" Quellenstudien zur
Musik: Wolfgang Schmieder zum 70. Geburtstag. Edited by Kurt Dorfmiiller, 69-77.
Frankfurt: C.F. Peters, 1972.

Hubbard, Frank. Three Centuries of Harpsichord Making. Cambridge: Harvard
University Press, 1965.

Kilian, Dietrich. Kritischer Bericht to NBA VII/3, Konzerte fitr Violine, fiir zwei
Violinen, fiir Cembalo, Flote und Violine. 1989.

Klein, Hans-Gfinter. Der Einfluss der Vivaldischen Konzertform im Instrumentalwerk
Johann Sebastian Bachs. Collection d ’Etudes Musicologiques; Sammlung
Musikwissenschaftlicher Abhandlungen, 54. Strasbourg: Editions P.H. Heitz, 1970.

Kobayashi, Yoshitake. ”Zur Chronologie der Spatwerke Johann Sebastian Bachs
Kompositions- und Auffiihrungstatigkeit von 1736 bis 1750.” Bach Jahrbuch 74
(1988): 7-72.

Kohler, Karl-Heinz. ”Zur Problematik der Violinsonaten mit obligatem Cembalo."
Bach Jahrbuch 45 (1958): 114-122.

Mackerness, E. D. "Bach’s F Major Violin Sonata." The Music Review 11 no. 3
(1950): 175-179.

Marissen, Michael. ”A Critical Reappraisal of J. S. Bach’s A Major Flute Sonata. " The
Journal of Musicology 6 (1988): 367-386.

337

------- . ”A Trio in C major for recorder, violin and continuo by J .S. Bach?" Early
Music 13 (1985): 384-390.

Marshall, Robert. ”Bach the Progressive: Observations on his Later Works." The
Musical Quarterly 62 (1976): 313-357.

------- . "J.S. Bach’s Compositions for Solo Flute: A Reconsideration of the Authenticity
and Chronology. " Journal of the American Musicological Society 32 (1979): 463-497.

Mattheson, Johann. Das beschiitze Orchestre. Hamburg: 1717.

------- . Das Neu-Erbfi‘nete Orchestre. Hamburg, 1713. Facsimile reprint Hildesheim:
Georg Olms Verlag, 1993.

------- . Der vollkommene Capellmeister. Revised Translation by Ernest C. Harriss.
Ann Arbor: UMI Research Press, 1981.

Mersmann, Hans. Die Kammermusik. Vol. 3. Leipzig: Breitkopf und Hartel, 1933.

Moser, Hans Joachim. "J.S. Bachs Sechs Sonaten ffir Cembalo und Violine."
Zeitschn'ftfitr Musik 105 (1938): 1220-1223.

Engel, Karl. "Konzert," in Die Musik in Geschichte und Gegenwart, Allgemeine
Enzyklopadie der Musik, edited by Friedrich Blume. VII col. 1572. Kassel and Basel:
Barenreiter, 1958.

Neumann, Werner. ”Das ’Bachische Collegium Musicum’. " Bach Jahrbuch 67 (1960):
5-27.

Neumann, Werner, and Hans-Joachim Schulze, eds. Bach Dokumente. Vol. 1,
Schnftstiicke von der Hand Johann Sebastian Bachs. Kassel: Barenreiter, 1963.

------- . Bach Dokumente. Vol. 2, Fremdschnfiliche und gedruckte Dokumente zur
Lebensbeschichte Johann Sebastian Bachs 1685-1750. Kassel: Barenreiter, 1969.

Neupert, Hanns. Harpsichord Manual. Kassel: Barenreiter, 1968.

Newman, William. ”A Checklist of The Earliest Keyboard ’Sonatas,’ 1641-1738.”
Notes 11 (1954): 201-212.

------- . ”Concerning the Accompanied Clavier Sonata." The Musical Quarterly 33
(1947): 327-349.

------- . The Sonata in the Baroque Era. Chapel Hill: The University of North Carolina
Press, 1959.

338

------ . The Sonata in the Classic Era. Chapel Hill: The University of North Carolina
Press, 1963.

Pook, Wilfrid. ”Bach’s E Major Violin Concerto Reconsidered." Music and Letters 38
(1957): 53-65.

Rees, Abraham. Universal Dictionary of Arts, Sciences, and Literature. London: A.
Strahan, 1819.

Rifkin, Joshua. ”Ein langsamer Konzertsatz Johann Sebastian Bachs. " Bach Jahrbuch
64 (1978): 140-147.

Rowen, Ruth Halle. Early Chamber Music. New York: Columbia University Press,
1949.

Scheibe, Johann Adolphe. Der Critischer Musikus: Neue, vermehrte und verbesserte
Auflage. Leipzig: Breitkopf, 1745.

Schering, Arnold. Musikgeschichte Leipzigs. Vol. 3. Leipzig: F. Kistner & C. F.
Siegel, 1941.

Schmieder, Wolfgang. Thematisch-systernatisches Verzeichnis der musikalischen Werke
von Johann Sebastian Bach. Bach-Werke Verzeichnis (BWV). Leipzig: Breitkopf und
Hartel, 1971.

Schmitz, Hans-Peter. Querfldte und Querflotenspiel. Kassel: Barenreiter, 1952.
------- . Kritischer Bericht to NBA VI/3, Werke fur Flb'te. 1963.

Schulze, Hans-Joachim and Christoph Wolff, eds. Bach Compendium: Analytisch-
bibliographisches Repertorium der Werke Johann Sebastian Bachs. Leipzig: Edition
Peters, 1985-.

Schulze, Hans-Joachim. ”Johann Sebastian Bachs Konzertbearbeitungen nach Vivaldi
und anderen: Studien oder Auftragswerke?” Deutsches Jahrbuch der
Musikwissenschaft 18 (1978): 80-100.

------- . "Johann Sebastian Bachs Konzerte: Fragen der Uberlieferung und Chronologie. "
Beitrage zum Konzertschafi'en Johann Sebastian Bachs. Bach-Studien 6. Edited by
Peter Ahnsehl, Karl Heller, und Hans-Joachim Schulze, 9-26. Leipzig: Breitkopf und
Harte], 1981.

------- . Preface to Konzert D—dur fur Cembalo und Streichorchester, BWV I 054.
Faksimile-Reihe Bachischer Werke und Schriftstucke, 11. Leipzig: VEB Deutscher
Verlag fiir Musik, 1972.

339

Shanet, H. ”Why did J .S. Bach Transpose his Arrangements?” The Musical Quarterly
36 (1950): 180-203.

Sheldon, David A. ”The Galant Syle Revisited and Re-evaluated." Acta Musicologica
43 (1975): 240-270.

------- . ”The Transition from Trio to Cembalo-Obbligato Sonata in the Works of J. G.
and C. H. Graun.” Journal of the American Musicological Society 24 (1971): 393-
413.

Siegele, Ulrich. Kompositionsweise und Bearbeitungstechnik in der Instrumentalmusik
Johann Sebastian Bachs. Neuhausen-Stuttgart: Hanssler-Verlag, 1975.

Spiro, Friedrich. ”Ein verlorenes Werk Johann Sebastian Bachs. " Zeitschrift der
Intemationalen Musikgesellschaft 7 (1905): 100-104.

Spitta, Philipp. Johann Sebastian Bach. 3 Vols. 1899. Reprint (3 vols. in 2), New
York: Dover Publications, 1951.

Stauffer, George. ”Bach as Reviser of his own Keyboard Works.” Early Music 13
(1985): 185-198.

Stephan, Rudolf. "Die Wandlung der Konzertform bei Bach." Musikforschung 6
(1953): 127-143.

Stinson, Russell. ”Toward a Chronology of Bach’s Instrumental Music: Observations
on 3 Keyboard Works. " Journal of Musicology 7 (1989): 440-470.

Terry, C. Sanford. ”Bach’s Dresden Appointment." The Musical Times 73 (April,
1932): 315-316.

Uldall, Hans. ”Beitrage zur Frfihgeschichte des Klavierkonzerts. " Zeitschriftfiir
Musikwissenschaft 10 (1927-28): 139—152.

Vogt, Hans. Johann Sebastian Bach ’s Chamber Music: Background, Analyses,
Individual Works. Translated by Kenn Johnson. Portland: Amadeus Press, 1981.

Wade, Rachel. The Keyboard Concertos of C. P. E. Bach. Ann Arbor: UMI Research
Press, 1981.

Walther, Johann Gottfried. Musicalisches Lexicon. Leipzig: Wolfgang Deer, 1732.

Wierichs, Alfred. Die Sonatefiir obligates T asteninstrument und Violine bis zum Beginn
der Hochklassik in Deutschland. Kassel: Barenreiter, 1981.

340

Williams, Peter. "Bach’s G Minor Sonata for viola da gamba and harpsichord BWV
1029: A seventh Brandenburg Concerto?” Early Music 12 (1984): 345-354.

Wolff, Christoph. "Bach’s Leipzig Chamber Music.” Early Music 13 (1985): 165-175.

------- . ”Johann Sebastian Bachs Spatwerk: Versuch einer Definition. " In J. S. Bachs
Spa'twerk und dessen Unfeld: Prospektiven und Probleme. Edited by Christoph Wolff,
15-22. Kassel: Barenreiter, 1988.

Wolff, Christoph, et al. The New Grove Bach Family. New York: W. W. Norton and
Company, 1983.

 

"111111111111111“