-
Music & Letters,Vol. 88 No. 3, The Author (2007). Published
by Oxford University Press. All rights
reserved.doi:10.1093/ml/gcm003, available online at
www.ml.oxfordjournals.org
BEETHOVEN AND THE DOUBLE BAR
BY BARRYCOOPER*
WHERE IN HIS MUSIC DID BEETHOVEN put double bars, and what
design did he use forthem? These might seem naive questions, for
which the answer could surely be found inany good modern edition of
his music. The questions might equally be viewed asabstruse and
esoteric ones of no conceivable practical interest. Both these
assumptionsprove to be erroneous, however, and an investigation of
Beethovens double bars unveilsfar more issues than might initially
have been expected.Any modern edition of Beethovens music, indeed
of most music, shows final double
barsthose at the end of a movement or workas a thin line
followed by a thick line,and intermediate double bars as two thin
lines. Before a repeated section there are dotspreceded by a
thick^thin pair of lines (often omitted if it is the start of a
work), matchedby dots and a concluding thick^thin pair of lines at
the end of it. These conventions areso familiar that they are
hardly worth recording, and they have been almost univer-sally
adopted by publishers for well over a century. Many musicians,
faced with suchconsistency, will no doubt assume that the same
conventions were observed inBeethovens day. This is not the case,
however. They date back much less than twocenturies, and what
happened before then has not been properly documented.
Somecomposers, copyists, and engravers used more ornate double bars
than those commontoday, and certainly Beethovens specimens cannot
be expected to look the same in hisautograph scores as they do in
modern editions. A glance at any of these autographscores quickly
confirms that the shape of his double bars differs from
modernconvention.If the design of his double bars is less than
straightforward, the actual location of
themother than those at the ends of movements or worksis an even
bigger pro-blem, despite the best efforts of recent scholarly
editors. Many recent editions ofBeethovens music, especially those
that describe themselves as urtext, make a pointof indicating how
carefully they preserve the composers original text. They
suggestthat this text is altered only in a few specified
situations, where editorial material isclearly indicated by such
means as square brackets or small notes. A notable example isthe
new complete edition currently being produced under the auspices of
theBeethovenhaus, Bonn, in which the more recent volumes (after
1991) include ErnstHerttrichs statement of the latest editorial
policy, containing the following passage:The edition reproduces the
music text largely in line with the notational style in
theauthentic sources. Characteristic scribal idiosyncrasies of
Beethoven are retained evenif they initially appear unfamiliar to
the reader. Modernizations are undertaken only if
*University of Manchester. Email:
[email protected]
458
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
they are necessary for the better readability and understanding
of the text. But they arenot allowed to alter its sense.1
Herttrich explains the various ways in which Beethovens text
might be altered, butsays nothing about double bars, leaving
readers to assume that they have not beenchanged, since any
addition might conceivably change the sense of the music.
Despitethis professed editorial policy, however, double bars are
routinely inserted at all changesof key signature or time
signature, and occasionally elsewhere. This has been doneapparently
so as to conform to the house style of G. Henle, the publishers,
withoutany evident recognition that it contradicts Beethovens
text.A similar situation occurs in Jonathan Del Mars justly
acclaimed and meticulously
prepared edition of Beethovens symphonies, published by Ba
renreiter. Here the editorstates in each study score: Wherever
possible, Beethovens own notation, nomenclature,clefs, spelling of
dynamic and tempo markings, and note-groupings have beenretained.2
Since it would be perfectly possible to retain Beethovens original
barring,which forms an integral part of his own notation, the
reader might well assume thatthis had been done here.Yet it has
not. In the score of the Ninth Symphony, the Prefacestates: It was
not Beethovens practice to write double bars at tempo or key
changes, butthis modern convention has been allowed.3 Thus in the
finale of this symphony, forexample, the edition includes
twenty-four internal double bars, none of which is in theautograph.
The question of whether this might make a difference to the
readersperception of the music is not considered, and the
distinction between places whereBeethoven did and did not use an
internal double bar becomes obscured.The intention to retain
Beethovens original notation is less explicitly articulated in
Peter Hauschilds recent edition of the piano sonatas, published
by Schott and Universalin their Wiener Urtext Edition, but the
edition still gives every impression that thisnotation is being
faithfully reproduced. Even the staccato signs are printed in
anunusual shape that attempts to mimic the composers handwriting,
and the prefacestates: Occasional editorial additions to the
musical text are indicated by squarebrackets.4 Once again, however,
double bars are added liberallywithout squarebracketsat most
changes of tempo or key signature, even if the music is
continuous.Thus three important and recent scholarly critical
editions of Beethovens music
prepared by different editors and issued by different publishers
have all disregardedthis aspect of his notation. These editions are
typical of most others in this respect. Attimes it seems as if
editors and publishers have not even noticed what was being done
toBeethovens notation, since they rarely mention double bars.
Meanwhile any user ofthese editions would be likely to assume,
wrongly, that the modern conventions fordouble bars were just as
stable in Beethovens day, and that no changes had been made.If the
double bars in modern editions differed from those in Beethovens
autographs
merely through insertions of them at routine changes of key
signature or time signaturewithin a movement, as implied by Del
Mar, such insertions would often cause no
1 Die Ausgabe gibt den Notentext weitgehend entsprechend der
Notierungsweise in den authentischen Quellenwieder.
Charakteristische, ausdrucksbedingte Schreibeigentu mlichkeiten
Beethovens, auch wenn sie dem Leserzuna chst ungewohnt erscheinen,
werden beibehalten. Modernisierungen werden nur dann vorgenommen,
wenn siezur besseren Lesbarkeit undVersta ndlichket des Textes
notwendig sind. Sie du rfen aber dessen Sinn nicht vera ndern.See
e.g. Ludwig van Beethoven, Musik zu Egmont und andere Schauspielen,
ed. Helmut Hell, Beethoven Werke, ix,7 (Munich, 1998), p. ix.
2 See e.g. Ludwig van Beethoven, Symphony 9 in D minor (study
score), ed. Jonathan Del Mar (Kassel, 1999), p. vi.The wording
varies slightly in the other eight scores.
3 Ibid.4 Ludwig van Beethoven, Sonaten fu r Klavier, ed. Peter
Hauschild, 3 vols. (Vienna, 1997^2001), i, p. viii.
459
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
problem, and might even help by drawing attention to a change
that could otherwisebe overlooked. Sometimes, however, they can
affect the interpretation of the music.A good example occurs at bar
91 of the finale of the Ninth Symphony. Here Beethovenwrote a
strong perfect cadence, followed by a single bar-line and then
immediately thefirst note of the famous Freude theme.Yet at this
point many conductors make a pause,often of several seconds, and it
seems highly probable that the presence of a spuriousdouble bar
here in modern editions has influenced this decision.
Interestingly, JamesWebster has already argued on purely analytical
grounds that there is no true caesurahere, and that this cadence
leads (or should lead) without break or loss of momentuminto the
Ode to Joy.5
Another problem is that many of Beethovens works were composed
in sections ormovements that are not quite independent but join on
to what follows in some way thatvaries from work to work, thus
offering a variety of boundaries with different levels ofstructural
significance. In the Waldstein Sonata, Op. 53, for instance, the
final Rondois preceded by an Introduzione that ends with a pause on
the dominant in preparationfor the Rondo theme. The theoretical
question of whether Introduzione and Rondoform two separate
movementsa slow movement and a finaleor a single movementlike the
slow introduction and Allegro in a first movement, is complemented
by thepractical question of whether there should be a momentary
silence between the two orwhether the first note of the Rondo
should be struck as the last note of the Introduzioneis released.
The same sort of issue is raised in many of Beethovens other works,
such asthe Appassionata Sonata, the Violin Concerto, and the
Emperor Concerto, where,although the music is not completely
continuous, the finale follows the end of theprevious movement
without a perfect cadence, usually with the word attacca to
confirmthat little or no break is envisaged. In the Fifth and Sixth
Symphonies, meanwhile, thefirst chord of the finale actually
concludes a phrase at the end of the previous move-ment, so that no
break is really possible, whether or not Beethoven wrote a double
bar.A slightly different situation occurs at the end of the first
movement of the MoonlightSonata, where there is a firm perfect
cadence in the tonic but Beethoven has added thewords attacca
subito il seguente. A particularly celebrated case where the
movementsare mostly run together, but to differing degrees, is the
String Quartet in C sharpMinor, Op. 131, with its seven movements
or sections. In works such as these, shouldthere be an additional
beat or two between movements, where possible, or should therhythm
continue uninterrupted? Sets of variations also raise the same
issue. Should theunderlying pulse continue smoothly from one
variation to the next, as would be neces-sary if they were
accompanying a dance; or should there be a brief pause or silence
atthe end of each one, perhaps preceded by a slight rallentando and
emphasis on thecadence, as in consecutive verses of a strophic song
or hymn?The kind of double bar that Beethoven wrote at all these
structural boundaries, and
whether he used a double bar rather than a single bar-line (or
nothing at all), mightthrow light on these questions. It follows,
therefore, that the theoretical question ofwhat types of double bar
he habitually used in his manuscripts, either at the end of orin
the middle of movements, could in some cases have significant
implications forperformance. This raises further questions that
have not previously been addressed in
5 James Webster, The Form of the Finale of Beethovens Ninth
Symphony, Beethoven Forum, 1 (1992),25^62 at 46.Websters music
example (p. 47) actually shows the single bar-line, though he does
not use this as part ofhis argument.
460
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
the literature: how many different kinds of double-bar sign did
he use, and was theirselection on each occasion made thoughtfully
or randomly?Beethoven himself was aware of the problems that double
bars could generate.When
George Thomson sent him some folksong melodies to harmonize,
Beethoven noticedthat many had a double bar in the middle, which
might imply a repeat of each half.After hinting at the problem in a
couple of letters, he eventually asked explicitly forclarification:
Note down for me in each song . . .whether there are any
repetitions :k:which are sometimes very badly notated by these two
k lines.6 Thus he was clearlyaware that the precise design of the
double bar could have significant implications, andit is therefore
to be expected that he was careful about how he drew his own and
wherehe placed them. To understand what options he had for the
sign, however, it isnecessary to explore briefly its history and
use before his time.
DOUBLE BARS IN OTHER SOURCES
Double bars, despite their familiarity or perhaps because of it,
have hardly ever beenthe subject of scholarly investigation. In New
Grove II, the article on double bar saysmerely this: Two vertical
lines drawn through the staff to mark off a section of a
piece.7
Thus the entire history and usage of a sign that has been widely
employed for manycenturies is reduced to less than a sentence. A
detailed search through the literature,using a number of standard
search mechanisms, failed to reveal a single article or bookin
which the term double bar (or Doppelstrich) appeared in the title.A
few scholars, however, have used double bars as an integral part of
an investiga-
tion, and five recent cases are worthy of note. Michaela Zackova
Rossi has discussedthe use of signs such as the corona and the
double bar as separation signs in the ballataand early madrigal; an
article by EllenTeSelle Boal describes Purcells use of fermatasigns
and double bars to mark sections of a work; Paul Cienniwa has noted
howCouperins markings in binary-form movements have been replaced
in modern editionsby standard double bars with dots, which may be
misleading; Christine Martinexplains that the two sections of the
final chorus in Handels anthem O praise the Lordwith one consent
are not separated by a double bar in the autograph and should
beregarded as continuous; and Simon Perry has observed that
Musorgsky uses doublebars in his Gnomus to indicate new sections
where the notation of accidentals ischanged.8 Even these rare
studies, however, do not specifically focus on double barsbut
merely bring them into the general argument. In the context of
double-bar study,they are noteworthy in three main ways. They
reveal first that the investigation ofdouble bars can yield
important insights that cannot be achieved by other means.Second,
works in which double bars have some real significance are not
limited to asingle composer or period but can come from a wide
variety of periods and genres,
6 note s moi a chaque chanson . . . sil y a des re pe titions
:k: qui Sont quelquefois tre' s mal note par ces deux k
lignes.Letter of 29 Feb. 1812. See Ludwig van Beethoven:
Briefwechsel Gesamtausgabe, ed. Sieghard Brandenburg, 7 vols.
(Munich,1996^8), ii, item no. 556.
7 See New Grove II, vii. 519. A cross-reference to the article
on baroffers nothing fresh, apart from giving the phraseup to the
double baras one example of how the word bar can be used.
8 Michaela Zackova Rossi, Ballata Form in the Early Madrigal, in
Anne-Emanuelle Ceulemans andBonnie J. Blackburn (eds.),The orie et
analyse musicales: 1450^1650 (Louvain-la-Neuve, 2001), 149^93;
Ellen TeSelle Boal,Tempo Indications in Purcells Fantasias and
Sonatas: A Performers Guide to New and Conflicting
Signatures,Journal of the Viola da Gamba Society of America, 31
(1994), 9^24; Paul Cienniwa, Repeat Signs and Binary Form
inFrancois Couperins Pie' ces de clavecin, Early Music, 30 (2002),
94^103; Christine Martin, Zur Form desSchlusschores von Ha ndels
Chandos-Anthem O praise the Lord with one consent (HWV 254), Go
ttingerHa ndel-Beitra ge, 7 (1998), 179^81; Simon Perry,
Mussorgskys Gnomus: Composers Score as Analytical Text,
Context,15^16 (1998), 5^20.
461
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
ranging from the medieval ballata to Musorgsky. And third, these
studies indicate thatany problems arising from double bars have
been overlooked in the vast proportion ofthe musical repertory.
Thus these writers demonstrate that the issue of double barsoffers
much opportunity for exploration.The double bar does have a long
and fairly complex history. It originated around the
end of the thirteenth century, as an extension of the Franconian
idea of a single bar-lineor finis punctorum, which had been
introduced by Franco of Cologne (fl. 1280) to indicatethe end of a
piece or section. In the Ars Nova period, a double or even triple
bar-linewas sometimes used to show the conclusion of an individual
voice-part.9 The implica-tion, perhaps, was that the space between
the two lines of the double bar represented aperiod of silence that
must occur at the end of every piece before the start of
anotherone. Since that time, the sign has taken on a variety of
forms and uses. Sometimes itindicated a repetition of a section, as
Beethoven mentioned in his letter to Thomson,and in such cases it
was often, though not always, accompanied by two or more dotsbefore
and/or after the sign. This usage can be seen, for example, in the
virginalcollection Parthenia (1612^13), where repeats are indicated
by two short vertical linessurrounded by dots on both sides, which
are in turn surrounded by two more verticallines. In this
collection, the ends of pieces that do not finish with a repeat
sign aremarked by a line of near-vertical zigzags on each stave,
gradually diminishing in heightfrom left to right, and finishing
with a final flourish. Such a sign is not untypical ofmany scores
of that period. Other sources use two plain vertical lines, thin or
thick,sometimes followed by further lines of smaller height. In
Book I of Bachs Daswohltemperirte Clavier (1722), each prelude and
fugue ends with two plain vertical lineson each stave, though Bachs
pen stroke sometimes joins them together and occasionallyextends
them with an additional shorter vertical line or flourish, thus
tending towardsthe zigzag design in Parthenia. In addition, he
places a zigzag line (of gradually dimin-ishing height) further to
the right, between the two staves, at the end of each fuguewhere
there is room, but only at the end of about half the preludes. At
the end of thelast fugue this zigzag is much larger and bolder, and
is followed by the word Fine.Thushe is not wholly consistent, and
further investigation in his other manuscripts is needed.In Mozarts
Requiem a rather different picture emerges. The opening Introit
ends witha half-close followed by a plain double bar (two long
vertical lines), but the Kyrie fugueends with a single bar-line
(divided in the middle), decorated by four pause-marks atvarious
heights and followed by two w-shaped squiggles at different heights
thatemphasize a stronger conclusion. The Dies irae, like the
Introit, also ends with a plainlong double bar before the Tuba
mirum; but later the Confutatis movement ends withjust a single
long bar-line and segue before the Lacrymosa.10 Su ssmayrs
contribution isnotated differently: he concludes movements with
short and often rather ornate doublebars, usually one on each stave
though sometimes covering 2^4 staves, and each with afinal diagonal
line up to the right.Thus sources exhibit considerable variation on
the basic design, which is two vertical
lines that may be joined together and may be followed by shorter
ones, often with somekind of final flourish, and perhaps with some
other kind of decoration. A full survey ofthe range of designs and
uses of double bars obviously cannot be attempted here,
9 See Richard Rastall,The Notation ofWestern Music (London,
1983), 55 and 62.10 Even in Christoph Wolff s diplomatic
transcription of the manuscript, this last detail is not preserved
and a
thin-thin double bar is substituted: see ChristophWolff, Mozarts
Requiem: Historical and Analytical StudiesDocumentsScore, trans.
MaryWhittall (Oxford, 1994), 217.
462
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
but the above examples show that Beethoven was faced with a
number of optionswhen he started writing music down and having to
make a conscious decisionabout how to indicate the end of a work,
movement, or section. His solution in individ-ual cases varied
considerably, but a survey of a large selection of his
autographsprovides a clear picture of his usual procedures. Once
this picture has been established,it can be used as a yardstick
against which to measure anomalous examples. It can alsobe used as
a means of illuminating the meaning of individual double bars (or
theabsence of them) in particular works, with results that are
sometimes surprising andsignificant.
BEETHOVENS NORMAL DOUBLE-BAR DESIGNS
The present survey of Beethovens double bars has been greatly
assisted by tworesources that have recently become available: one
is the microfiche publication of allthe Beethoven manuscripts in
the Berlin Staatsbibliothek (including those currently inKrako w);
the other is the online digital archive of the sources in the
Beethoven-Archiv,Bonn,11 which enables anyone with Internet access
to view numerous examples ofBeethovens different designs of double
bars. Between them these two collections con-tain a large majority
of his extant autograph scores, but some additional autographshave
also been examined to amplify the picture. The only previous
extended study ofthe subject appears to be by Douglas Johnson,12
who is one of the few scholars to havediscussed Beethovens double
bars at all. Johnson constructed an impressive list of thefinal
double bars in all Beethovens autographs up to 1798, in an attempt
to use them(along with other notational features) as an aid to
dating the early works.The principlewas perfectly sound and worked
well with certain handwriting features, although withdouble bars it
proved in the end more a matter of using the works, dated by
othermeans, as a way of dating Beethovens use of particular
designs. Precisely how thesedesigns were used when they appeared in
the middle of works was not part of Johnsonsinvestigation, and the
categories he used for classifying the different types of double
barhave needed to be adjusted for the present study, for he does
not always distinguishbetween different versions of a general
design.Altogether Beethoven used four main designs of double bar
during the course of his
life (see Pl. 1). The first consisted of a pair of vertical
lines on each stave of the system,with each pair superimposed by a
reverse s-shape. This type (Johnsons type C) will bereferred to
hereafter as the s-type. Since Beethoven almost invariably drew
each of hissingle bar-lines across all the staves in a systemwith a
single stroke of the pen, the use ofan individual sign like this on
each stave is quite striking visually. This sign is usuallyfound in
roughly the same places as the modern thin^thick double bar, and
was hismost common conclusion sign for whole movements and complete
works up to 1801. Itfirst appeared as early as 1786 in his Trio for
piano, flute, and bassoon (WoO 37),13 butat this stage it
invariably took on some irregularity: in the first movement, a
single longpair of vertical lines was combined with a reverse-s on
each stave; the second move-ment concludes with a different type of
double bar (see below); and in the third, a set
11 Musikhandschriften der Staatsbibliothek zu BerlinPreuischer
Kulturbesitz, Teil 3: Die Beethoven-Sammlung [microfichesand
accompanying booklets] (Munich, 2002^5).The Bonn sources are
accessible at5www.beethoven-haus-bonn.de4.
12 Douglas Johnson, Beethovens Early Sketches in the Fischhof
Miscellany: Berlin, Autograph 28, 2 vols. (Ann Arbor,1980),esp. i.
39^41, 60^3.
13 For the location of this and other autographs discussed, see
the Appendix.
463
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
of variations, the s-type is used only in combination with dots
signifying repeats(Theme, Var. 1where the dots follow the s-typeand
Vars. 3, 4, and 6).14 The workends with an irregular type, a set of
four long vertical lines superimposed by a series ofloops that form
a coil. The s-type was used regularly during the 1790s, however,
andJohnson lists seventeen works up to 1799, in addition toWoO 37,
where it can be found.It continued in common use in autographs
around 1800, as in the finales of the con-certos Opp. 15 and 19,
the first, third, and fourth movements of the Piano Sonata, Op.26,
the G major Violin Romance, Op. 40, and the finales of the Piano
Sonata Op. 28and String Quintet Op. 29. At this stage, however, it
was less common at the ends ofnon-final movements, such as those of
Opp. 15, 19, 28, and 29, where other types areused instead.The last
regular use of the s-type may be in the six Gellert Lieder, Op.
48,which were probably composed not long before a copy of them that
is dated 8 March1802. The surviving portion of the autograph
contains only the last two songs, with no.6 appearing in two
versions, but the s-type appears consistently in all three
settings. Bythe time Beethoven came to write out the autographs of
his Three Violin Sonatas,Op. 30, however, around spring 1802, he
was no longer using the s-type at the ends ofmovements, and from
this date onwards any use of this type must be regarded asirregular
and deserving special attention.One particularly interesting
anomaly is theThird Piano Concerto, Op. 37. The date
of this work has long been a matter of dispute, for there is
evidence that it was largelycomposed in 1800, in preparation for a
benefit concert that spring, but it was not finallyperformed until
April 1803. Virtually all the sketches are missing, and the date on
theautograph appears to be 1803, though it was for a long time
misread as 1800. LeonPlantinga has argued that the most likely date
for the missing sketches is summer 1802,and that none of the score
was written out until early 1803;15 but there is insufficient gapin
the sketching record to allow for the composition of a large-scale
orchestral work insummer 1802, for Beethoven was composing three
piano sonatas (Op. 31) at that time,having interrupted two sets of
variations (Opp. 34 and 35) to complete these quickly.Indeed any
proposed date in 1802 or 1803 is problematical in this respect, for
thesketching record shows no major gaps. The double bars offer some
dating evidence notpreviously considered: both the second and third
movements of Op. 37 end with thes-type. It is extremely unlikely
that Beethoven would revert to this type for this onework in early
1803, having abandoned it so thoroughly and consistently almost a
yearearlier. Thus the concerto must have been written out in score
by early 1802 at the
PL. 1. Beethovens notation of double bars
14 Johnson asserts that the s-type is not normally used with
dots as a repeat sign (Beethovens Early Sketches, i. 40), butthere
are actually quite a few cases in addition to WoO 37, viz. Op. 103,
3rd mvt., WoO 1, 6th mvt., WoO 33, no. 5,WoO 52, and Hess 48.
15 See Leon Plantinga, Beethovens Concertos: History, Style,
Performance (New York and London, 1999), 113^35, wherethere is a
detailed discussion of the dating problems of Op. 37.
464
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
latestmore likely 1800 or 1801with the final bar reached at this
stage, whateveradditions and amendments were made later (and there
were many). This conclusion is,incidentally, strengthened by the
form of system brace used in the score. In 1800Beethoven was
already moving to his most mature form of system brace, in
whichthere is no separate arc at the top, and a mixture of this and
earlier forms can befound in the Piano Concerto in C, the present
score of which dates from 1800.16 By 1801the form with the arc was
rarely used, as can be seen in such autographs as Opp. 27 No.2, 28,
29, 40, andWoO 46 from that year,17 and by 1802 it hardly ever
appears; yet it canbe found almost throughout the score of the C
minor Piano Concerto. It seemsimpossible to believe that Beethoven
would suddenly and briefly return to his olderform of system
braceand his older form of double barfor a single work, and
theconsistency with which the system brace with separate arc is
used suggests early 1801 asthe latest possible date for the main
outline of the score, with 1800 more probable. Thedate 1803 on the
score must therefore have been added by Beethoven only whenmaking
late revisions around the time of its premiere that year. His
double bars,together with his system braces, therefore provide
crucial evidence for dating one ofhis major works.The second type
of double bar found in Beethovens autographs corresponds
approximately, like the s-type, to the modern thin^thick double
bar, and is describedby Johnson as a single stroke, beginning with
a vertical line which spans one staffand continuing to the right as
a squiggle of gradually diminishing height.18 Thedesign most often
has three, but sometimes two or four, vertical strokes
alternatingwith oblique ones, and can often resemble a roughly
drawn letter m; it will thereforebe referred to here as Beethovens
m-type of double bar. Like the s-type, it was drawnseparately for
each stave of a system. It was used for a short time in 1794,
appearing in apiece for musical clock (WoO 33 No. 4), the revised
score of Der freie Mann (WoO 117),and the first two movements of
Beethovens copy of Haydns String Quartet Op. 20 No.1, all of which
appear to date from that year.19 There is also a puzzling use of
the m-type in a work dating from c.1791theVariationsWoO 67; an
explanation can be foundfor this anomaly, however, for the sign
appears only in a correction added on the titlepage.20 This
correction was surely made in 1794, for Beethoven wrote to
NikolausSimrock in Bonn on 18 June that year, offering to send the
score of this work andadding that various improvements had been
made in this score.21 Simrock publishedthe work later that year.The
m-type was quickly abandoned that same year, but it began to
reappear around
the end of 1801. Its reappearance was not entirely sudden, for
there are a few transi-tional specimens from that date. One of the
earliest is the first movement of the PianoSonata in D, Op. 28,
where the sign on each stave has two downstrokes linked by
anoblique, and thus resembles a short plain double bar. A similar
form appears at the endof the second movement of the String Quintet
Op. 29, of roughly the same date.
16 Johnson, Beethovens Early Sketches, i. 37, 358. The work was
originally composed five years earlier.17 It does appear frequently
in the solo part of Op. 19 written out in April of that year; but
this manuscript was a
fair copy intended for a printer, and Beethoven habitually
tended to use an older, more formal style of writing infair copies
than in composing scores; see e.g. Johnson, Beethovens Early
Sketches, i. 34^5.
18 Ibid. 40. He classifies this sign as type E.19 A further
example occurs in a sketch in the Kafka Miscellany, fo. 141v: see
Ludwig van Beethoven: Autograph
Miscellany from circa 1786 to 1799, ed. Joseph Kerman, 2 vols.
(London, 1970). Johnson dates this leaf as 1794 on othergrounds
(Beethovens Early Sketches, i. 102), and the m-type of double bar
supports his conclusion.
20 Johnson, Beethovens Early Sketches, i. 484 n. 10.21
Briefwechsel, ed. Brandenburg, i, item no. 15.
465
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
Both these works show the s-type at the end of the finale, but
the one in Op. 29 couldalso be considered transitional, since the
two vertical strokes are somewhat linkedtogether, thus creating
something of a hybrid between the m-type and s-type. Beforelong,
however, the fully-fledged m-type was appearing regularly, and can
be found atthe end of each of the ten movements in the Violin
Sonatas of Op. 30, apart from theScherzo of No. 2, which ends with
repeat signs. It also appears at the end of each of theseven
bagatelles, Op. 33, and the two sets of variations, Opp. 34 and 35,
which all datefrom later that year.Thus the transition from regular
s-type to regular m-type seems tohave lasted around six months,
with Opp. 28 and 29 showing both types, Op. 48showing only the
s-type and Op. 30 No. 1, written about April 1802, showing only
them-type. Thereafter one finds the m-type at the end of
practically every work, and ofnearly all movements that end with a
firm tonic cadence. This feature remained con-sistent throughout
the rest of Beethovens output, and the very few exceptions appear
tohave special circumstances (as with the Third Piano Concerto,
discussed earlier). Them-type does not normally appear, however, at
the end of inconclusive movements orsections such as those
mentioned earlier, where a different sign is used. The
inescapableconclusion, therefore, is that from about April 1802
Beethoven normally treated them-type as signifying and confirming a
strong closure. It can even appear occasionallyat a major
structural point within a movement, where there is a firm tonic
cadenceand change of metre. An example occurs in the long final
movement of Christusam Oelberge, Op. 85, in a revision apparently
made in 1804.22 Here the initial recitativeends at bar 25 with a
firm perfect cadence before the ensuingTerzetto, and
Beethovenunexpectedly uses an m-type. This suggests a very strong
termination, and that asubstantial pause was envisaged before the
start of the Terzettoproviding a goodexample of how the use of a
particular double-bar sign can affect the way a work isperceived
and performed. In a modern edition, a thin-thick double bar would
beequivalent, and should surely be used here.In a few late works,
Beethoven saved time by writing the m-type double bar only on
the first-violin stave.This procedure can be seen, for example,
in the Kyrie of theMissasolemnis, Op. 123, and the first movement
of the Ninth Symphony. It never became hisnormal habit, however,
and there does not appear to be any musical significance in
thisvariant of the sign.Beethovens third type of commonly used
double bar is his repeat sign, consisting of a
pair of short vertical lines on each stave, preceded and/or
followed by dots, dependingon which passage was to be repeated (as
in modern notation).This sign could appear atthe end of a movement
instead of one of the other types, or in the middle in place of
abar-line, or even between two bar-lines in the middle of a bar. It
was sometimesdecorated by two pairs of slanting slashes, one above
and one below the stave. Since itprovides no clues on its own about
the division of a work into sections or movements, ithas no real
structural significance and can therefore be excluded from the
followingdiscussion. If a repeat occurred at the end of a movement,
the dots of the repeat signwere occasionally combined with the
s-type of double bar, as mentioned above, or withthe m-type, but
the latter combination was very rare (see below for its use in Opp.
34and 35).The fourth type of double bar frequently found in
Beethovens autographs is the
plainest: two long vertical lines, each drawn through all the
staves of a system
22 See AlanTyson,The 1803 Version of Beethovens Christus am
Oelberge, Musical Quarterly, 56 (1970), 551^84, esp. 568and 578.
The original 1803 autograph is missing.
466
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
(Johnsons type A). Beethoven used this extensively in his
sketches throughout his life, tomark the end of a section or
movement, and it proved a good way of distinguishingsuch sketches
from the majority that needed some continuation. Sometimes it was
alsoused to mark the beginning or end of some fragment, and it is
the most common type ofdouble bar in his sketches, although other
types are occasionally found. In finishedscores, however, it is
much less common, and is used almost invariably to denote
aninternal division or structural boundary, signifying that there
was more to come. Thusit is somewhat like the modern thin^thin
double bar, but was used less often. Johnsonlists only three early
works where the sign appears at the end, and all are special
cases.One is the very early set of three piano quartets, WoO 36
(1785), written beforeBeethoven had established his normal
patterns; here he used a short pair of verticallines on each stave
at the end of each movement, adding dots for repeats
wherenecessary. Another is in the Duo, WoO 32 (c.1797), where again
there are pairs ofshort vertical lines on each stave, but only at
the end of the Minuet; since this isfollowed by a Trio, it is not
really a proper ending (the Trio itself ends with a repeatsign).
The third is the end of theVariations,WoO 74, where there was very
little spacefor any bar-line, let alone a decorated double bar, and
Beethoven drew two squashedand superimposed long bar-lines, with an
additional squiggle to suggest a trueconclusion.In contrast to
these three special cases, Beethovens normal use of the double bar
as a
sign of incompleteness emerged as early as c.1786, where it can
be found in the fragmen-tary orchestral Romance, Hess 13, and the
Trio,WoO 37. In the Romance, he unchar-acteristically drew all his
bar-lines on single staves rather than using long ones across
allthe staves, and the same applies to his double bars, which he
added to mark the start ofa solo section and again before a
maggiore section. InWoO 37, the slow movement endson an imperfect
cadence, and Beethoven added just a long plain double bar,
followedby two linking chords and rests, then only a single
bar-line before the start of the finale.Thus in both these works he
already demonstrates a perception of a plain double bar asa
non-ending, unlike in the earlierWoO 36 trios.This perception
continued in later life,and typical examples of the use of this
sign at incomplete endings include theIntroduzione in the Waldstein
Sonata; the Allegretto section in the finale of the samesonata,
immediately before the final Prestissimo; individual variations
inWoO 74; theCoda that precedes the Finale in the
PrometheusVariations, Op. 35; the third movementin the oratorio
Christus am Oelberge; the slow movement of the Violin Concerto, Op.
61,before Attacca subito il Rondo; and the third movement of the
Fifth Symphony, whichhas the sign at the end of both the minore
section and the whole movement. The signalso appears in many other
works in similar contexts. In the String Quintet Op. 29,when the
music changes to 3/4 towards the end of the finale, Beethoven
unusually put ashort double bar on each of the five staves, but no
further double bars are used at thesubsequent time changes. The
same sign was used about the same time in theContredanse,WoO 14 No.
10, at the end of both the main section and the Trio section.Here
Beethoven probably considered that neither section was a proper
conclusion, sincethere was a da capo to follow, and the two
sections would be played in alternation for aslong as was required
by the dance that the music was intended to accompany; a
plaindouble bar was therefore appropriate. (The other contredanses
written at this time,nos. 9, 7, and 2, all use repeat
signs.)Beethovens use of this fourth type of double bar is the most
variable and interesting,
and his habit of reserving it mainly for sketches and for
incomplete endings has impor-tant implications for places where it
appears unexpectedly. There are several of these,
467
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
and they need special consideration. Its use in sketches seems
to have extended beyondtrue sketches to some complete compositions
that he evidently did not regard as properworks. Thus they appear
at the end of both formal counterpoint studies23 and
othercompositional exercises. These include the Fugue in C of
c.1795 (Hess 64), a shortAdagio for three horns (Hess 297) written
in 1815, and even the substantial recitativeand aria No, non
turbarti (WoO 92a). This last work was written around the
beginningof 1802 during Beethovens study of Italian word-setting
with Salieri, and the fact thathe did not regard it as a proper
composition is evident not only from Salieris correc-tions but from
Beethovens own heading: Esercizi. The final plain double bar
reinforcesthis view.One place where a plain long double bar appears
unexpectedly is the end of the
second movement of the C major Piano Concerto. Although the word
attaccadoes notappear, the use of this double bar rather than a
proper termination sign may havesignified that the finale should
continue with very little break.The sign would thus haveserved as a
visual reminder when Beethoven was performing and conducting from
thisscore. There would be no problem in omitting a break at this
point, since the thirdmovement begins with piano solo and the horns
have time to change crooks before theyenter.24 One could of course
suspect that Beethoven was merely being inattentive orcasual about
what kind of double bar he put here, but for the fact that he used
exactlythe same sign in the corresponding place in his next
concerto, the Third PianoConcerto. Here he initially wrote a plain
long double bar, but then drafted the lasttwo bars again, using
conventional s-type signs, as noted above; the supplementarybars,
together with the s-type signs, were later cancelled, leaving the
plain long doublebar at the end of the movement. This was clearly
not casual or through inattention butwas a conscious decision that
must have had some meaning. Either he was experiment-ing with using
a different type of double bar for non-final movements, or more
probablyhe intended the sign as a reminder to proceed without delay
to the solo passage thatbegins the finale. Again this would cause
no problems for the horns, for they do notchange crooks at this
point (and the trumpets and timpani have been silent during theslow
movement). The musical effect of continuing promptly here is to
emphasizethe connection between the G sharp at the end of the slow
movement and the A flatat the opening of the finalean enharmonic
relationship that has intrigued severalcommentators. The first
movement of this concerto also ends with a plain double bar,and so
again it must be supposed that a prompt continuation may have been
envisaged.Many more double bars also appear during the course of
this work, but these weresimply to mark off the solo sections from
the tutti ones, and seem to have been addedto aid the copyist,
though they do not appear with complete consistency.25 Their usein
similar places can also be found in the fragmentaryViolin Concerto
in C,WoO 5,of c.1791, and in the first movement of the Piano
Concerto in B Flat, but these areexceptional cases.A plain double
bar can also be found at the end of a middle movement in the
Piano
Sonata Op. 28, second and third movements. At the end of the
third movement this isjustified since there is a da capo, but there
seems no obvious musical reason why the
23 Facsimiles of some examples can be seen in Alfred Mann,
Beethovens Contrapuntal Studies with Haydn,Musical Quarterly, 56
(1970), 711^26, pl. I^III.
24 The time allowed for a change of crook in bars 1^20 is almost
exactly the same as in the finale of theThird PianoConcerto, bars
229^48.
25 In the first movement they appear after bars 111, 130, 171,
179, 226, 257, 322, and 347, but not e.g. 402.
468
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
second should proceed to the third any more rapidly than usual.
This may be whatBeethoven wanted, however; alternatively he may
simply have been experimentingwith a sign to differentiate a middle
movement from a final one: a reminder to con-tinue, but not
necessarily with an attacca. This unusual case occurred during the
periodof his transition from s-type to m-type, when some
experimentation might be expected.A particularly notable use of a
plain double bar is in the Violin Romance in F,
Op. 50. The score of this was written out in 1798, but there are
several reasons forsuspecting that the work may be a revision of
the missing slow movement of the violinconcertoWoO 5, of which only
a fragment of the first movement survives: the Romancehas the right
title, key, speed, scoring, and theme for this purpose,26 and
independentinstrumental Romances were scarcely known at that time,
although many slowmovements had that title. Moreover there are no
sketches for Op. 50 in Beethovenssketchbook of 1798, as there
should have been if the work were newly composed.The heading in the
autograph provides a further clue, for it looks like that of a
middlemovement, with no proper work title and no composers name.
What really seemsto clinch the matter, however, is the use of a
plain long double bar at the end, just asin the slow movements of
several other concertos (Opp. 15, 37, 61), for among all
theautograph scores examined, no other examples of this sign have
been found at the endof a self-contained work that was not an
exercise. The one apparent exception is theSixth Symphony, but here
the last bar, including the double-bar sign, is actually inthe hand
of a copyist, the original page having disappeared. Thus the double
bar inOp. 50 evidently signifies that there must have been a
movement to follow. The Op. 40Romance in G, by contrast, has a
standard s-type of conclusion sign, as already noted.It must be
deduced, therefore, that Beethoven revised his early C major violin
concerto,or part of it, in 1798, then extracted the Romance for a
separate performance; and thatthis won so much admiration that some
patron commissioned another Romance, whichBeethoven duly provided
by composing Op. 40 as an independent work.One other case of a
plain double bar at the end of a movement is worth noting. This
occurs with the first movement of the Piano Sonata Op. 109. Here
Beethoven originallywrote his customary m-type, but immediately
smudged it out and replaced it with aplain double bar on each
stave, plus an attacca. As Nicholas Marston has pointed out,What
Beethoven did at the end of the first movement of Op. 109 was to
replace thisdouble barline [the m-type] with the more provisional
form consisting of two ordinarybarlines placed close together. . ..
That he substituted this form for the more conven-tional one
clearly indicates that he imputed different meanings to the two
signs.27 Theheightened continuity was then confirmed by a pedal
mark on the last chord of the firstmovement, and matching pedal-off
sign on the first chord of the next movement,ensuring there was no
silence between the two (hence Beethoven felt free to delete
theattacca that he had added). Beethoven also used a plain long
double bar at the end of thesecond movement. This implies that the
whole sonata is more unified and continuousthan usual.28 In the two
companion sonatas, however, Beethoven used firm m-typedouble bars
at the end of all three movements in Op. 110 and both movements
inOp. 111. Thus those pianists (and there are a few) who make no
break between move-ments in either of these sonatas are not playing
what Beethoven wrote. Although theend of the second movement of Op.
110, for example, with its prolonged F major chord
26 See Barry Cooper, Beethoven (Oxford, 2000), 75^6.27 Nicholas
Marston, Beethovens Piano Sonata in E, Op. 109 (Oxford, 1995),
10.28 See ibid. 9^11, where some of the analytical implications of
the plain double bars are developed.
469
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
and widespread sonority, neatly prepares for the next movement,
the m-type doublebar is reinforced by the absence of any attacca,
confirming that Beethoven did notintend immediate continuity (the
m-type seems never to have been used in conjunctionwith an
attacca). In any case, the close tonal relationship cannot be used
as an argu-ment, for it resembles that between the end of Op. 10
No. 1 and the beginning of Op. 10No. 2, which one would not expect
to be played without a break.
IRREGULARTYPES
Apart from his four main types of double-bar sign, Beethoven
used a few others, thoughonly irregularly and mainly in his early
works. The use of a set of short plain doublebars each covering
just a single stave has already been mentioned, as has a coil of
loopssuperimposed on four long vertical lines. In what is probably
the earliest autograph ofall, the Fugue WoO 31 of c.1783, there is
a unique ending with pairs of vertical linesadorned by horizontal
rather than vertical zigzags (Johnsons type F), but the coiledloops
at the end of WoO 37 could be seen as an elaboration of the
horizontal zigzags.Further examples of a coil of loops, combined
with two long vertical lines, appear insome scores of c.1790^2: the
first movement of the Duo for Flutes, WoO 26; certainvariations in
the set for piano, WoO 64; and the song with orchestra Mit Ma deln
sichvertragen, WoO 90. The loops, however, could easily degenerate
when Beethoven waswriting more hurriedly, leaving just a descending
wavy line (usually about four waves)over the two straight vertical
lines. This type is found at the end of the first movementof the
Ritterballett of 1791 (WoO1), although the other movements all have
the ordinarys-type (as do most of those in the autograph piano
reduction). It also appears in somevariations inWoO 64, and in two
works for voice and orchestra from about the samedate: Pru fung des
Ku ssens and Primo amore (WoO 89 and 92). Occasionally
Beethovencombined waves and loops within a single design: this
occurs in the second and fourthmovements of the Wind Octet, Op.
103, of 1793, where in both cases the two longvertical lines are
superimposed by a wavy line in which the waves become looped inits
lower half. This suggests that he regarded waves and loops as
interchangeable at thisperiod. Both types virtually disappear after
1793, though a rare later example,with only two to three waves, can
be found in the first movement of the PianoConcerto in C (1800).A
different design found during the 1790s is a pair of long vertical
lines decorated by
a single elongated reverse-s; hence it can be referred to as a
long S-type.29 This type,usually drawn quite neatly, is uncommon,
and its function is rather variable. Often itdenotes a specifically
non-final ending, as in the following: the sketchy accompanimentfor
the Lamentations of Jeremiah (c.1791), where the sign appears at
the end of the firstverse but not most of the later verses; the
trio section of No. 6 in the piano score of theRitterballett, where
it is followed by a da capo; one variation in the set on La' ci
darem,WoO 28 (c.1795); two variations in the first movement of the
sonata Op. 26; and the firstmovement of the Moonlight Sonata, where
the sign is accompanied by the wordattacca. The sign also appears
in the score of the Piano Concerto in B flat at the endof the first
two movements, whereas the finale uses the short s-type. In these
works,therefore, the symbol distinguishes some intermediate endings
from the final one.
29 Johnson (Beethovens Early Sketches, i. 40) places this in the
same category as double bars with loops or waves,claiming that the
backwards S could become a long wavy line when more than two or
three staves were involved.These categories are better treated
separately, however, for they are not very similar, and the wave
type can also befound with only two staves (e.g.WoO 64), while the
long S-type sometimes appears across more than three staves.
470
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
Occasionally, however, the reverse is more the case. In the
piano score of theRitterballett, where most movements end with the
short s-type, the finale ends with thelong S-type, decorated with a
pause and the word fine. A nearly complete fantasia-likepiano piece
in D (c.1793) also ends with the long S-type, after intermediate
sections hadended with plain double bars; and the symbol is used
for all three movements in thesolo part of the Piano Concerto in B
flat, Op. 19, which was written out in April 1801 togo with the
orchestral score completed earlier. Nevertheless, the sign appears
to be usedat the end of a complete work only in piano reductions
(e.g. of Op. 19) and sketchyworks (e.g. the piece in D) where some
continuation might have been planned. Henceits general meaning
seems to be that the score is in some way less than complete at
thatpoint. The latest use of the sign yet discovered occurs in an
incomplete attempt atconcluding the slow movement of the Second
Razumovsky Quartet (1806). WhenBeethoven rewrote the final bars he
substituted his normal m-type of double bar. Thusthe long S-type
here may have been a reminder to himself that the first attempt
wasonly a provisional draft to be replaced. If so, the sign again
signalled that the work wasless than complete, but in a different
sense than usual.Since Beethoven sometimes stretched his
single-stave s-type across several staves to
form a long S-type, it is not surprising that he also extended
his m-type across morethan one stave.This is infrequent, however,
and appears mainly in the sketchbooks (e.g.Landsberg 5, p. 50;
Wittgenstein, fo. 41v), normally on just two staves. It can also
befound at the end of the set of bagatelles Op.126, whereas the
rest of the bagatelles in thisset, and those in the previous set
(Op. 119 Nos. 1^6), end with the usual m-type on eachstave. In this
case, therefore, Beethoven was treating the two-stave m-type as
more finalthan the one-stave m-type.
EXCEPTIONAL ANDUNEXPECTEDUSAGES
In general, then, Beethoven usually used an s-type or m-type
double bar, but occasion-ally some other type, for the end of a
work or a movement with a clear perfect cadence,and he used a plain
double bar for inconclusive endings, where something else was
tofollow promptly if not immediately. Exceptions to this pattern
deserve greater scrutiny,since there may well be some underlying
intention that caused the irregularity. By farthe most common
exception is the complete absence of a double bar between
closelyadjoining sections of a work. A double bar was omitted not
only at temporary changesof key or metre, as in the finale of the
Ninth Symphony (see above), but sometimes atmore substantial
boundaries.The absence of any double bar in such places
subliminallyconveys to the performer a much greater sense of
continuity, and offers much lesstemptation to slow down or make a
slight break for a boundary in the music. Thissituation occurs, for
instance, before several finales. Although Beethoven used a
plaindouble bar before the finale in works such as the Waldstein
Sonata, the ViolinConcerto in D, the Fifth Symphony, and the third
Razumovsky Quartet, he avoidedone altogether in certain other
works, using just a single bar-line immediately beforethe start of
the finale, as in the Appassionata Sonata, the Sixth Symphony, and
theFifth Piano Concerto.Is there any reason for differentiating
these three finales from the others just men-
tioned? Musical features suggest that there is, and that
Beethoven was being moresystematic than might be expected. In the
Appassionata, the opening diminished-seventh chords of the finale
are prefigured in the last two bars of the precedingAndante, so
that in a sense the finale has already begun before the change of
keysignature and time signature. Moreover the pedal at the end of
the Andante is probably
471
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
intended to be held right through (as in the first two movements
of Op. 109); althoughthere is a fresh pedal mark at the start of
the finale, it is probably to be interpreted assempre pedale, since
the previous pedal mark two bars earlier has no cancellation. In
theFifth Piano Concerto the situation is somewhat similar. The main
finale theme isprefigured in the last two bars of the Adagio, the
modulation to the original tonichaving already taken place, and so
once again the finale has effectively already begunbefore the
change of metre. Thus a double bar would be less appropriate.
Indeed thisprinciple was already foreshadowed in the very
earlyTrioWoO 37, as noted earlier.TheSixth Symphony is somewhat
different. It is always regarded today as having fivemovements, but
the fourth, the Sturm, is shorter and less well developed than
theothers, and lies outside the traditional four-movement
structure. Its structural prece-dents are introductions to finales
such as those in Beethovens Septet and his First andThird
Symphonies, and especially La Malinconia in the Quartet Op. 18 No.
6, and heseems to have been building on this tradition. Although
the Sturm is longer than theseintroductions, with a clear musical
narrative, it is only slightly longer than the slowintroduction at
the start of the Seventh Symphony (about seventeen seconds longer,
ifone observes Beethovens metronome marks); this introduction also
has a clear musicalnarrative but has never been considered a
separate movement. The single bar-line atthe end of the Sturm (in
contrast to the plain double bar at the end of the
precedingmovement, Lustiges Zusammensein der Landleute) is
therefore appropriate, and itstrengthens the view of the Sturm as
in some ways a prolonged introduction to thefinale rather than
simply a full-scale movement in its own right, thus providing
aninteresting alternative perspective to the works overall
structure. It is certainly anoversimplification to suggest that the
Sixth Symphony has five proper movements butthe Seventh only four,
as is commonly asserted.There are many other places where only a
single bar-line is used at a transition
between sections of a work. Early examples include the joins
between the march andthe trio (and back to the march) in the Marcia
funebremovement of the Piano SonataOp. 26, and that between the
orchestral introduction and opening recitative in Christusam
Oelberge. Several other cases occur in the bagatelles Op. 33. If
such a transition to anew section took place during a bar,
Beethoven wrote no bar-line at all but just a freshset of key or
time signatures. Similarly, at the end of the Scherzo of Op. 26,
where theTrio ends on the second beat of the bar, there is no
bar-line after the last note, but just ada capo instruction. In the
multi-sectional Fantasia Op. 77, there is not one double
barthroughout the piece until the customary m-type at the end. This
results in some veryodd bar lengths and unexpected absences of
bar-lines, especially near the end, wherethere is no bar-line
immediately before the tempo 1mo mark, where the main themereturns
in C major.In the multi-sectional works of Beethovens later years,
there are further places in
which modern editions customarily add a double bar where there
is none in the auto-graph. In the Piano Sonata Op. 101 there is no
bar-line after the Adagio movementbefore theTempo del primo pezzo,
and just a single bar-line between this and the finalAllegro. In
the Cello Sonata Op. 102 No. 1, a single bar-line is drawn between
theAdagio and Tempo d Andante, and again between this and the
ensuing Allegrovivace. The end of the first movement is more
complicated: Beethoven put a plainlong double bar and attacca
indication, but he cancelled the latter in the copyistsscore; an
attacca would anyway have been somewhat weakened by the final bar
of thefirst movement, which is a whole-bar rest with pause. Had he
rewritten the autograph,he might well have replaced his plain
double bar with an m-type at this point.
472
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
The companion sonata, No. 2 in D, has no bar-line at the end of
the second movement,but just Atacca [sic] and a fresh set of
signatures (clef, key, and time). Thus the lastchord of the Adagio
and first note of the Allegro are part of the same bar in
theautograph score. Similarly, in the multi-sectional finale of the
Piano Sonata Op. 110there are no internal double bars, and changes
of key or time within a bar take placewithout any added bar-line;
hence the first note of the Fuga falls within the same bar asthe
last chord of the preceding Adagio.30
One of the most extraordinary scores from this point of view is
the String Quartet inC Sharp Minor, Op. 131. As mentioned earlier,
this has seven movements which adjoinin different ways. At the end
of the first movement there is no attacca sign, butBeethoven links
the two movements even more closely together: there is not even
asingle bar-line between them, and the first chord of the second
movement comes in thesame bar as the last chord of the first
movement. Although the page shows a large spacebefore the second
movement, a fresh set of clefs that might imply some sort of break
hasbeen deleted. A similar level of continuity occurs at the end of
the next two movements,although they both happen to terminate at
the end of a bar, and so a single bar-lineseparates them from what
follows. At the end of the fourth movement, however,Beethoven wrote
a firm m-type conclusion sign, the type used at the end of
wholeworks, thus indicating a very strong caesura before the rest
of the quartet. (There aretwo autograph scores for this part of the
work, and in the later one he wrote the m-typeonly on the
first-violin stave, as in the first movement of the Ninth Symphony,
but thishas no evident difference in meaning.) The fifth and sixth
movements once again endwith only a single bar-line in the
autograph, although Beethoven drew a long plaindouble bar two bars
before the end of the fifth movement, signifying that this
move-ment really ends here and that the following two bars form a
link to what follows. Thiswas not a new procedure, for it had
appeared as early asWoO 37 (c.1786), as alreadynoted.The links
between movements therefore differ from what one would deduce
from
modern scores, which add a spurious double bar after each of the
first three move-ments, creating a stronger division, but imply
some continuity at the end of the fourthmovement instead of a
full-scale break.The biggest surprise in the autograph, however,is
the end of the finale: there is no bar-line or other conclusion
sign at all! DidBeethoven really just forget to put one in, and
never notice the glaring omissionwhen going through his score? This
seems unlikely. Was he uncertain about whetherto add an additional
eighth movement that appears in his sketches, and planned toinsert
whatever bar-line or double bar was appropriate once he had reached
a decision?One would still expect him to put in some kind of sign,
and then change it later ifnecessary, as in other cases. Whatever
the explanation, the absence of any double barbeautifully matches
the unstable and ambiguous ending in the music itself, where theC
sharp major chord can be heard either as a strong tonic or as a
weak dominant of Fsharp minor that had been heard a few bars
earlier, and Beethoven may have beenresponding almost instinctively
to this inconclusiveness.Apart from the m-type sign at the end of
the fourth movement, and the long
double bar just before the end of the fifth, the only double-bar
signs in thisquartet appear during the fourth movement, at the end
of the theme and each ofthe first four variations, though not after
the fifth or sixth. There is no break in the
30 In this movement, for once, Hauschilds edition (see n. 4
above) mostly follows the autograph score, adding veryfew redundant
double bars.
473
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
flow of the music at these points, and Beethoven may have
decided to insert thesedouble bars (which are not in his earlier
autograph) merely to help orientate theplayers, and perhaps to
encourage them to strengthen the sense of cadence at each ofthese
five points. It is quite noteworthy, however, that the variations
are more separatedfrom each other visually by double bars than any
of the first four movements or thelast three.
OTHER SETS OF VARIATIONS
In the variation movement in Op. 131, musical continuity between
variations isgenerated by the use of short note values on weak
beats and half-beats at the end ofthe theme and each variation, and
Beethoven used either plain double bars or singlebar-lines. But
what did he use in his earlier sets of variations, where each
section(theme or variation) often ends with a strong cadence and
long note or rest? In somecases the section ends with a repeat
sign, giving little clue about continuity since, as hasbeen seen,
this sign can appear equally at the end of a movement or in the
middle ofa bar with no difference in design; but where there is no
repeat at the end, the typeof double bar used might provide clues
about how much continuity he appears to haveenvisaged between
variations. One of the earliest sets of variations surviving
inautograph score is WoO 64. Here, the theme and each variation end
with a firmcadence and no link to what follows; as already noted,
at each point Beethoven drewa long double bar decorated by either
loops or waves. All but two of these doublebarsthose for the theme
and Var. 2are also decorated by a pause. Although suchpauses do not
appear in later sets of variations, they imply that he considered a
slightbreak between variations to be normal, as with consecutive
verses of a song. It ispossible, therefore, that he wanted less
break or even no break after the theme andVar. 2, where there is no
pause. Although there is no obvious reason why this shouldhappen,
it would be quite effective. The theme ends with a long three-beat
note, andone might want to press on to Var. 1 with little or no
delay after this. Var. 3 is theminor-key variation, and so again
one might want to avoid delay at the end of Var. 2,so as to enhance
the sudden change of character. Conversely, at the end of the
ratherserious minor variation a slightly longer pause than usual
would suit its contemplativecharacter well. Thus it could be
significant that Beethoven uses a different type ofdouble bar here
than before. This and the final double bar are the only two that
usewaves rather than loops, which might mean that he imagined a
stronger terminationhere than elsewhere, although the change of
design could of course be just a randomdeviation.Some similarly
irregular double bars, which could be interpreted as either
deliberate
or random deviations, occur in several later sets of variations.
In the woodwindvariations on La' ci darem (WoO 28) of c.1795, the
theme and all but one of thevariations have the short s-type of
double bar (if one regards the coda as belongingwith the final
variation), and they are normally followed by a space on the page.
Theexception is Var. 3, where Beethoven used the long S-type, and
left no space after itbefore Var. 4. Perhaps, therefore, he was
using a different type of double bar to implygreater continuity
than after the other variations, as implied by the absence of a
space.It may also be significant that no new time signature is
written at the start of Var. 4(though this is also true of Vars. 2,
6, and 7), which might suggest closer continuity. Themusic does
cadence properly at the end of Var. 3, butVar. 4 is the first to
begin with anupbeat before the main downbeat, and this could have
prompted him to use a different
474
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
type of double bar and no following space, to emphasize the
greater continuityrequired.Some sets of variations from the late
1790s show no significant pattern of double-bar
design; for example the variations WoO 74 all use long plain
double bars, while thevariation movement in the Septet Op. 20 uses
an appropriate mixture of the s-type(which implies a break between
variations) and repeat signs. In the Piano SonataOp. 26, however,
we find a pattern that recalls those inWoO 64 andWoO 28. At theend
of theTheme, as at the end of Vars. 1, 3, and 5 (if one counts the
short coda as partof Var. 5), Beethoven used his normal s-type,
implying a firm ending plus short silence.Yet at the ends of Vars.
2 and 4 he used the long S-type, which was normally reservedfor
non-final or provisional endings, as established above. At
precisely the same twoboundaries he also omitted the time signature
that he normally included in eachvariation in a setVars. 3 and 5
are the only two that lack one. Thus the double barsand the time
signatures concur in implying that he may have wanted a shorter
breakhere than at the end of theTheme or the other variations.
There is no obvious musicalcontinuity at the ends of Vars. 2 and 4,
but it may be significant that his pupil CarlCzerny later singled
these two variations out as the two that should be
performedslightly faster than the rest.31 If this reflects
Beethovens own view, as is very possible,then it would make sense
for them also to lead on more quickly to what follows, in linewith
the implications of the double bars and time signatures. It is also
noteworthy that,as in WoO 64, the implication is for a shorter
break before the minore variation anda longer one after it.By the
time Beethoven composed his next set of variations within a sonata,
in the
finale of theViolin Sonata in A (Op. 30 No. 1) of 1802, his
normal conclusion sign wasthe m-type of double bar. This sign
appears not only at the end of each movement, butalso at the end of
each variation, except Vars. 1 and 5. Thus one must assume
asignificant break after most of the variations. Since Var. 5 ends
with a pause on thedominant, it is not surprising to find just a
plain double bar, before the change to 6/8 forVar. 6 and coda, and
the implication is that Var. 6 should follow almost immediately.The
end of Var. 1 is more intriguing, for Beethoven used the short
s-type on all threestaves, and it is followed by the only variation
for which he did not write a fresh timesignature. More
significantly,Var. 1 is the only variation, apart fromVar. 5, that
ends onthe weak fourth beat, thereby inviting a more immediate
continuation. Thus, as inOp. 26, a different kind of double bar is
combined with an absence of following timesignature and greater
musical continuity with what follows. In Op. 30 No. 1, however,the
m-type had supplanted the s-type as the standard conclusion sign,
releasing thelatter so that it could imply a less conclusive ending
than formerly, while the longS-type could be abandoned altogether.
Thus the consistency with which Beethovenvaries his double-bar
signs in several different sets of variations, including WoO 64,WoO
28, Op. 26, and Op. 30 No. 1, strongly suggests that the
distribution of irregularsigns is not merely random but relates to
the music itself, with implications for how suchvariations might be
performed.Later in 1802 Beethoven composed two major new sets of
variations for piano,
Opp. 34 and 35. Op. 34 is highly innovative in that each
variation is followed by achange of key and time signature. The
theme and Var. 1 both end with an m-type ofdouble bar, but the next
three variations all end with repeats. Beethoven here created
31 Carl Czerny, On the Proper Performance of all BeethovensWorks
for the Piano, ed. Paul Badura-Skoda (Vienna,1970), 37.
475
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
an unusual combination of repeat sign and the m-type, as if to
signify that he wanted aproper pause at the end of each variation,
after the repeat. He originally put the samesign at the end of Var.
5, where there is a firm cadence in C minor, but he then deletedthe
sign and substituted an ordinary repeat sign, followed by a few
supplementary barsthat lead to a pause on a dominant seventh of F
major in preparation forVar. 6 (whichis preceded, as expected, by a
long double bar). This change confirms that his use ofrepeats
combined with an m-type signified something different from and
moreconclusive than an ordinary repeat sign, since he felt it
necessary to delete the signwhen adding a linking passage straight
after it.The unusual combination of m-type with repeat sign
reappears in Op. 35, the
so-called Eroica Variations or, more properly, the Prometheus
Variations. Here theinitial Basso del Tema ends with a normal
m-type, while the next three sections,a due, a tre, and a quattro,
end with conventional repeat signs. The following Tema,however,
ends with repeat plus m-type, suggesting that Beethoven wanted a
breakbefore the main numbered set of variations is begun.Variations
1^3, 5, and 7 concludewith ordinary repeat signs that might or
might not suggest a slight break, while Var. 4has a separate
second-time bar followed by an m-type double bar; a break here
wouldbe quite effective before the next variation, which begins
pianissimo. Variation 14 runsinto Var. 15, and consequently
concludes with a long plain double bar, but the othervariations all
end with an m-type (supplemented by repeats in one case). Thus a
slightbreak between these variations seems to be the intention. But
there is an anomaly:Var. 6 leads into Var. 7 with a run of
semiquavers and must obviously join on withouta break, yet there is
a firm m-type double bar. If Beethovens double bars have
anysignificance, this must be regarded as an oversight, for the
m-type elsewhere almostinvariably coincides with a strong ending
and implies at least a slight break.Thus the answer to whether
Beethovens variations should be played with or without
a break after each one depends on the precise context. A break
seems to have beenexpected more often than not, and the types of
double bar used provide additional cluesin some cases. This
interpretation applies in most of the later sets of variations
too.In the slow movement of the Appassionata Sonata, Op. 57, the
Theme ends withambiguous repeat signs but the first variation ends
with the m-type of double bar, andboth of the first two variations
are given a fresh set of clefs and signatures.Thus a slightbreak
before each of them is strongly implied by the autograph. The
remainingvariations, however, proceed without a break, and without
any further double bars.In the third movement of the Archduke Trio,
Op. 97, every variation follows onwithout any break in the music,
and there is only a single bar-line each time, with nofresh
signatures. Even the end of the movement concludes with only a
single bar-linesince, as with certain works already mentioned, the
first chord of the finale is alreadyheard before the change of time
signature. The autograph score of the DiabelliVariations, Op. 120,
is in private hands and inaccessible, but a single replacement
pagesurvives in Bonn; this shows the end of Var. 31, with no double
bar or bar-line at theend before the start of the fugal Var. 32.
The continuity here is also emphasized by anattacca sign.Probably
the most complex relationship between variations occurs in the
finale of the
Piano Sonata, Op. 109. The Theme ends with repeat signs, but
none of the variationsdoes; nor do any of them have a firm m-type
sign (except, of course, after the reprise ofthe theme at the very
end).Variations 1 and 4 end with a plain long double bar and ashort
space afterwards; Vars. 2, 3, and 5 end with a single bar-line and
no space,implying that Vars. 3, 4, and 6 must begin immediately.
This implication is supported
476
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
by the music, for Vars. 2, 3, and 5 all end on the final
quarter-beat of the bar, whereastheTheme andVars. 1 and 4 have an
implied caesura, either through a stable cadenceor (Var. 4) a short
rest. Musically the link is strongest between Vars. 3 and 4,
wherethere is a note tied across the bar-line, and this is the only
place where Beethoven didnot add fresh clefs and key signature for
the start of the new variation. He did actuallybegin writing them,
but promptly smudged them out, and so the decision was
clearlydeliberate. Thus the double bars at the end of Vars. 1 and
4, and perhaps the repeatsigns at the end of theTheme, may imply a
bigger break than occurs at the ends of theother variations.
Marston has reached the same conclusion (although his reasoning
isbased just on the double bars and not the space after certain
variations or the rhythmof their final bars); and he has gone on to
argue that this interpretation is also structu-rally plausible,
with Var. 1 somewhat isolated while Vars. 2^4 form a
connectedgroup.32 One must guard against too much sense of
isolation here, however. Thelargest separation between variations
is represented only by a plain long double bar,which is much less
strong than the m-type used in some other sets, and presumably
lessstrong than the various modified forms of quasi-final double
bar found in sets such asOp. 26 and Op. 30 No. 1.
THETRANSMISSION OF BEETHOVENS DOUBLE BARS
Once Beethovens music began being copied and printed, his
complex range and choiceof double bars was unlikely to be
reproduced exactly, and there are numerous placeswhere it has been
distorted. He seems either not to have noticed or not to
havemindedpartly, no doubt, because he had more important matters
to deal with, suchas wrong notes and missing accidentals. He knew,
too, that copyists and engraverstended to have their own
idiosyncratic designs for double bars and would not usuallydraw
them precisely the same way as he did anyway. In general, however,
Beethovenscopyists reproduced his notation rather closely. Their
job was to copy what they saw,however much they suspected an error,
and so if he put a single bar-line where onemight expect a double
bar, they tended to do the same. A good example occurs in
thecopyists score of the Cello Sonata, Op. 102 No. 1, where the
scribe, like Beethoven, putonly a single bar-line after the Adagio,
and again after the Tempo dAndanteunlikemodern editions, where a
double bar is inserted. Copyists were liable to change verybizarre
notation, however. In the next sonata (No. 2), where Beethoven had
left anempty space at the end of the second movement, signalling
that the music runs on intothe finale, the copyist could not resist
adding a double bar, which Beethoven leftunchanged when he
corrected the manuscript. The final exposed bar at the endof Op.
131 was also closed off in the copyists score by a conventional
conclusionsignthe same as the one he used at the end of the fourth
movement (and differentfrom the more provisional signs at the ends
of the other movements, which mostlymatch the autograph).In
contrast to copyists, many publishers in Beethovens day felt it
their duty not
merely to print the music supplied but to edit it in various
ways. This seems to havebeen true more for German publishers such
as Simrock of Bonn and Breitkopf & Ha rtelof Leipzig than for
some of theViennese publishers, but individual publishers also
hadtheir own idiosyncrasies and no generalizations about this are
absolutely firm. Changessuch as added slurs, revised titles, and
altered layout were not uncommon in the early
32 Marston, Beethovens Piano Sonata in E, 10^11 and 239.
477
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
nineteenth century, and double bars were among the signs likely
to be edited. Mostconventional music at that time did not change
key signature or time signature exceptat major structural points,
where a double bar was likely to occur anyway.Thus doublebars
probably became associated in many editors minds with such changes,
so thatbefore long these editors were liable to feel it necessary
to add a double bar wheneversuch a change occurred, so as to create
an additional visual aid for the performersperception of the music.
Hence the insertion of double bars by publishers at all changesof
key signature or time signature became an increasingly common
convention, whichsurvives to the present day. Beethovens music was
unusual for its period in being morelikely than most to change key
or time at a point that was not structurally significant,and this
could give rise to spurious double bars being inserted by
publishers withoutproper justification.These problems become
particularly acute in Beethovens later music. In Op. 109,
Adolph Schlesingers engraver added two double bars not in
Beethovens autograph inthe first movement. The connection between
the first two movements was alsocompletely lost, since the engraver
used a thin^thick double bar to conclude the firstmovement, and
omitted the pedal-off sign at the start of the second movement,
therebyimplying that the two movements were completely separate.
Meanwhile the subtlestructure of the finale suggested by the use of
single and double bars was obscured,with every variation ending
with either a thin^thick or thick^thin double bar(the latter type
appears at the ends of Vars. 3^5). Adolphs son Moritz was no
betterwhen he came to publish the sonata Op. 111. At bar 44 there
is a change of keysignature, and in the first printing the engraver
followed Beethovens manuscript inhaving just a single bar-line; yet
when the sonata was re-engraved by the same firmonly a few months
later this was replaced by a double bar, although
Beethovenscorrection list made no such request. This change well
illustrates how double barswere infiltrating new editions at that
time.Schotts So hnen, of Mainz, were equally ready to add spurious
double bars. In the
bagatelle Op. 126 No. 1, there is a short passage of only nine
bars in 2/4 time, with therest in 3/4, but the publishers felt
obliged to surround this passage with double bars oneither side,
although there are none in the autograph and the music is
completelycontinuous.33 They made similar insertions in their
original edition of the NinthSymphony, so that, of the twenty-four
double bars added to the finale in Del Marsedition noted above,
twenty-two are already in the Schotts edition.When engraving
theinstrumental parts for Op. 131 (plate no. 2628), Schotts also
inserted unauthorizedthin^thin double bars between most of the
movements; but they did at least retainBeethovens (and his
copyists) clear division between movements 4 and 5, printing
athick^thick double bar and indenting the start of movement 5. Yet
when they came toengrave the score a little later (plate no. 2692),
this boundary was wrongly treated thesame as the other movements,
with a thin^thin double bar and immediate continua-tion. This
change has probably contributed to the confusion in later editions,
andgreater continuity in many performances than Beethoven
envisaged. Of course, somemusicians may prefer to have no break
here, just as some may prefer a long break afterbar 91of the finale
of the Ninth Symphony, or no breaks between movements in Op.110(see
above); but personal preferences of this sort are a separate
issue.
33 Facsimiles of both autograph and first edition are in Ludwig
van Beethoven, Sechs Bagatellen fu r Klavier Op. 126,ed. Sieghard
Brandenburg, 2 vols. (Bonn, 1984).
478
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
The way Beethovens notation could become distorted is well
illustrated in the finaleof the sonata Op. 110. Here the end of the
first Arioso (bar 26) happens to occur at thefoot of a page in his
first autograph, but the bar, or rather half-bar, is left open at
theend, and the Fuga begins at the top of the next page with a
change of key signature andtime signature.When Beethoven copied
this into his second autograph (the one now inBonn), the boundary
happened to occur in the middle of a page, and so within a
singlebar one finds the end of the Arioso, changes of signatures, a
new title and the first noteof the fugue. The copyist, however,
used the first autograph when preparing a score forthe printer, and
on seeing an open-ended bar at the end of the Arioso he added a
plainlong double bar (this was the same copyist, Wenzel Rampl, who
added double bars toopen-ended bars in Op. 102 No. 2 and Op. 131).
In addition, he indented the start of thefugue on the next system,
thus giving the impression that this was a completely newmovement.
In the original edition, Moritz Schlesingers engraver adopted the
samelayout, strengthening the boundary still further by using a
thin^thick double bar atthe end of the Arioso. The boundary was
still not quite as strong as at the end of thethree main movements,
where a thick^thick double bar was printed, but it
neverthelesslooks fairly conclusive and seems to invite a
substantial break before the fugue, creatinga four-movement
structure instead of a three-movement one. As usual, Beethoven
madeno attempt to modify this layout when he corrected the proofs.
Surprisingly, however,most modern editions print this particular
boundary correctly without any bar-lines,as in the autograph
scores.Not all publishers treated Beethovens bar-lines in such
cavalier fashion. When the
first edition of the Appassionata Sonata was printed inVienna,
the engraver correctlyput only a single bar-line at the end of the
slow movement. Meanwhile, in theHammerklavier Sonata, the original
Viennese edition included not one double barthroughout (apart from
those at the ends of the four movements and at one repeatsign),
despite numerous changes of key and several changes of time. This
doubtlessreflected Beethovens own notation in the lost autograph.
Most modern editions havenot been so faithful here, with many
double bars inserted.The original London edition,prepared from a
now lost manuscript sent by Beethoven, also adds a few double
barsalways thick^thick (a common design at that period), and
normally at changes oftimebut not as many as most modern
editions.In his autograph scores Beethoven often used abbreviated
notation that he expected
publishers to print in full (as he occasionally indicated in his
letters). But he evidentlydid not expect them to add double bars in
a similar way, since extra ones are not foundin editions such as
the Viennese one of Op. 106. The omission of a double bar in
anautograph was therefore not made to save time but reflected
something about thenature of the music. On the other hand, the fact
that he was prepared to toleratealterations to double bars by his
copyists and publisherseven to the extent of allowinga movement to
be split in two by Schlesingerwithout ever commenting on them inhis
letters or correction lists, or correcting or changing them in
copyists scores orpublishers proofs (as far as is known) reveals
that he did not attach great significanceto them. Thus one should
be wary of becoming too obsessed by them, or insisting thateach
type should be reflected precisely in any performance. Double bars
cannotactually be played, or conveyed reliably to listeners, and
the amount of difference theycan make to any performance is
strictly limited.Nevertheless, Beethovens double bars formed an
integral part of his compositional
conception, and should therefore always be taken into account in
the preparation ofeditions and performances of his music, if we are
to come as close as possible to
479
at Harvard University on July 25, 2010
http://ml.oxfordjournals.org
Dow
nloaded from
-
understanding his musical thoughts. As Roger Sessions has said,
Understanding ofmusic, as relevant for the listener, means the
ability to receive its full message.34
If, as Sessions and others would aver, music involves
communication between composerand listener, then any clues to what
the composer was trying to communicate are worthuncovering; and the
precise form of notation in an autograph, including the types
ofdouble bars (as well as such matters as beaming and stem
direction), can be veryrevealing, for it throws light on how
Beethoven envisaged the music at the time whenhe was most deeply
involved in creating its final shape.35 Adding an
unauthorizedinternal double bar might easily affect the
implications of the music concerning itsphrase structure and
continuity, leading to subtly altered perceptions by performersand
resultant modifications of how they play the passage. The presence
or absenceof internal double bars, as well as their actual design,
clearly does impinge on thesense of the music.
BEETHOVENS SYSTEM AND ITS IMPLICATIONS
If Beethovens use of double bars had been entirely regular,
investigation of themwould have been straightforward. Equally, if
it had been completely random anddisorganized, investigation could
have proceeded little further than simply establishingthat this was
the case. Instead, his use of double bars, like his music itself,
is extremelycomplex and sometimes impossible to predict even after
study of related scores, with abewildering variety of patterns that
has necessitated extensive investigation. It is,however, very
largely consistent and systematic, once its complexities have
beenu