Understanding Classical and Early Romantic Dynamics 1750-1830
Bart van Oort
Royal Conservatory The Hague
Research Project 2012-2014
Index
0. INTRODUCTION
2
I. Early Dynamics
1. Often no dynamics at all
2. Character
3. What if there is an opening dynamic?
4. Interpreting lacking dynamics
5
9
14
16
II. Understanding Dynamics
1. The performer's domain
2. Understanding the musical language
18
20
III. Implied dynamics
1. Introduction
2. Dissonance and consonance
3. The correct performance of the slur
4. Register
5. Touch
23
25
27
31
39
IV. Interpreting dynamics
1. Tempo rubato
2. Forte markings
3. Counter accentuation
4. Messa di voce
5. Crescendo as rinforzando
6. Decrescendo vs. diminuendo in Schubert
41
43
46
49
53
61
V. The question of subito piano
1. The fortepiano's natural decay
2. Beethoven
3. Actual subito piano
4. Subito fortissimo
68
68
72
73
VI. Case Studies
1. Mozart Sonata K. 283, Adagio
2. Beethoven Sonata in D minor, op 31/2 'Tempest'
3. Schubert Fantasie in C major, D. 605a, ' Grazer Fantasie'
75
75
77
80
VII. Conclusion
83
Bibliography
84
1
INTRODUCTION1
Luckily, there is no consensus about how the works of Mozart and Beethoven should be
performed. Present-day performances range from rich and romantic to clear and taut: some
pianists choose the modern piano, whilst others choose the 18th or early 19th century
fortepiano. Both have their own unique charms, possibilities and limitations.
Mozart, Beethoven and Schubert (as well as Schumann, Chopin, or Debussy) worked within
the framework of the musical and technical possibilities of the instruments they knew, which
each had their own complete universe of sound, technique, and articulation. While for some
modern ears the perceived dynamical and coloristic limitations of the fortepiano are
detrimental to the musical experience, it is a given that these composers, unhindered by these
so-called 'limitations,' created their greatest works within these boundaries. Furthermore, the
very qualities which today are seen by some as limitations often served as sources of
inspiration for both the composers and their public. One might argue for the use of the modern
piano in works of one hundred to two hundred years ago because it is an instrument of our
time: a logical argument given the way in which we experience the great masterpieces of the
classical period in the 21st century. Mozart, for example, never played in a concert hall of
2000 or more seats, and in spite of the wonderful acoustical qualities that such halls can have,
his instrument can never satisfactorily fill this space with its limited volume.
“But if Beethoven had known the modern piano, he would certainly have preferred it to the
fortepiano of his time…” The assumption that every development also signifies progress
doesn’t hold true for the fortepiano. Indeed, the development from Mozart’s piano to the
modern piano was not necessarily a change for the better. Naturally, when the most important
criteria of the 21st century listener is that the tune must be glowing and the melody-element
dominant, the modern piano is unsurpassed. But composers from the classical and early
romantic periods wrote music that not only sings, but pleas, argues, entreats, languishes,
persuades, sighs, gnashes its teeth, shouts for joy, whispers... in short, music that must also
speak. (Have you ever wondered where the word ‘recital’ comes from?) This repertoire is
deeply rooted in the rhetoric of the baroque, and it is in this rhetorical style where the
instruments of the time are so convincing.
Beethoven, the innovator, was according to legend also an inspiration for some of the
innovations in piano building. While we might ask ourselves how much of his calls for a
louder, fuller piano tone had to do with his increasing deafness, it must be said that every note
he wrote is excellently realized on the fortepianos of his time. He was pragmatic in every
sense, therefore, the remark “If Beethoven had known the modern piano…” doesn’t hold
water: if Beethoven had known the modern piano, he would have without a doubt written
different music. And this hypothetical music of Beethoven, as well as the modern piano, don’t
fit in the classical Zeitgeist. History teaches us that we can not pass over any step: Beethoven
also did not. He wrote rhetorical music that was effortlessly recognized during his time, and
that was performed on instruments which were highly valued universally.
As a result, in our time the differences between historic and modern instruments caused
different stylistic conceptions between the modern and “historical” schools of playing. The
fortepiano creates a clear musical texture, a prerequisite for rhetoric, by means of a silvery
tone that doesn’t sound too long, a sharp onset, and a large palette of colors. All of these
characteristics make it easy to articulate – to speak the musical words. (Naturally, this goes
1
Adaptation of the introduction to my Geske Lecture, published by the Hixon-Lied College of Fine and
Performing Arts of the University of Nebraska at Lincoln, NE, in 2009.
2
not only for the fortepiano: this is the fundamental difference between all historical
instruments and their modern counterparts.) The modern piano tone is warmer and longer, and
sings through more – it requires no great imagination to see that the development of piano
tone follows on the heels of opera. Herein lies, if the difference between the historical and
modern-day piano could be summarized in one line, the crux of the matter: while the
fortepiano speaks, the modern piano sings. And now the question is, whether the singing
modern instruments are really the most logical choice for the realization of the rhetorical
music from the 18th and early 19th centuries.
The changes in musical style over the past two hundred years occurred simultaneously with
the changes in instrument building. In other words, the changing style influenced the ideas of
what instruments should do, while at the same time, developments in instrument building
created a stimulus for that style change. The developments in building proved to be
irreversible, and allowed the style to develop in only one direction. The most important
changes in fortepiano building took place between 1750 and 1850, and with those
developments, the style of the early classical period had become history. Therefore, the
question is whether we still know how to interpret the scores of Mozart and Beethoven. Many
issues with regard to rhythm, rubato, pedaling, tone colour, dynamics, and last but not least
the actual sound and pronunciation, we may not fully understand. In addressing these issues,
the fortepiano proves to be an extremely valuable tool.
To play classical works on a modern piano means making a translation of sorts. As in any
translation one loses part of the original meaning in the process. The modern player, like a
translator, faces two problems: relative unfamiliarity with the language because two centuries
of history have blurred the picture, and a modern instrument which ‘pronounces the musical
words’ differently. The first problem is unrelated to the instrument, for it is one of
understanding the musical language. We can try to solve the second problem by finding a
‘native speaker’ to pronounce the language for us; in this case, the eighteenth century
fortepiano. The fortepiano is now widely available. In this article, I will attempt to address
some lesser known aspects of the classical musical language.
The rhetorical style requires, besides a sense for “speaking the musical words” and an
instrument which enables one to do so, a particular freedom in realizing the score. Far from
taking Mozart’s and Beethoven’s notation as the complete and unalterable truth, we must start
from the assumption that some aspects of their style could not be notated. Treatises of the
time support this assumption: for instance, all treatises before 1800 agree that the application
of dynamics belongs to the domain of the performer. As a result, early scores show little or no
dynamics at all. The dynamics that are notated are almost without exception very general.
The melancholy progress of the musical phrase or the minimal dynamic differences between
dissonance and consonance in, for example, a Mozart Adagio (or even a Chopin Nocturne for
that matter) are so subtle that they can not even be notated. Herein lie the deepest utterances
of the composer’s as well as the pianist’s soul, and they cannot be fixed in conventional
notation.
Pianists of all times have, of course, instinctively understood this: besides dynamics, elements
such as timing, rhythm (as opposed to meter), tone colour, tempo, tempo rubato, pedaling,
balance between the hands, drive and direction, and subtle, speech-like accentuation, cannot
be adequately recorded in our musical notation. But the degree to which these unwritten
liberties have been exploited has diminished greatly in the two hundred years since Mozart
died: with the disappearance of the improvising performer, performance of the classical
3
repertoire has become decidedly less free. The advent of the Urtext Edition has given a major
blow to the artistic initiative of many performers: rather than starting from knowledge about
the (often hidden) meaning of the notation and the resulting intentions of the composers,
present day music students often seem to literally follow the score while boasting 'fidelity' to
the text. These interpreters erroneously take that what is written down for instructions, rather
than descriptions. Only after we know more or less precisely what the composer wrote
(through a manuscript, a first edition, an Urtext) the real work of the interpreter starts, namely
to find out what he meant.
I will attempt to set parameters for the use of one of these unwritten liberties, namely
dynamics.2 I will not try to give new rules; they do not exist in a rhetorical style where the
personal voice of the orator gives life and meaning to every statement made.
2
In this article I will not comment on accentuation markings, nor on dynamics which seem to imply the use of a
pedal (like ppp, which may indicate the use of the moderator pedal in early romantic works, in particular in
Beethoven and Schubert). These could be the subject of extensive further study.
4
I. Early Dynamics
1. Often no dynamics at all
In much of the early keyboard repertory until ca 1800 hardly any, or no dynamics at all are
notated. This is the case in the earliest keyboard works which are definitely composed for the
harpsichord. This has always been considered logical, the harpsichord not being a touchsensitive instrument.
J.S. Bach. Partita no 1, BWV 825
But even for harpsichords with two manuals (on which loud and soft can be realized) one
hardly ever finds dynamic markings.
J.S. Bach. French Suite no 1, BWV 812
And, in fact, one does not even find dynamics in scores for instruments that are highly
dynamical:
5
J.S. Bach. Flute Partita BWV 1013 in A minor
J. S. Bach. Cello Suite no 1 in G, BWV 1007 - Prélude
Bach-Gezellschaft Ausgage, Bd 27/1. Leipzig: Breitkopf & Härtel, 1879
The most striking example is offered by the two part Inventions and three part Sinfonias,
written by JS Bach for the clavichord. Bach specifically mentions how he hopes, through
these works, that the students will achieve a singing quality of playing: 'am allermeisten aber
eine cantabile Art im Spielen zu erlangen'. Yet Bach does not notate any dynamics.
J. S. Bach. Inventions and Sinfonias.
H.G.M. Darnköhler, copyist / Manuscript, n.d.(1745-1755)
One wonders whether this lack of description of the intended dynamics has to do with a
harpsichord tradition. Haydn never indicated dynamics in his keyboard sonatas until the six
Auenbrugger Sonatas, Hob XVI/35-39 and 20, published in 1780. His early keyboard works
6
are by some thought to have been written for the harpsichord, but Haydn would have
definitely also have written them for and performed them on clavichords. Whatever the type
of keyboard, it did not seem to necessitate the addition of dynamic markings to the score.
Besides that, the limited dynamics of the harpsichord cannot by itself account for the missing
dynamical notation as there is a similar lack of indications in string quartets, symphonies and
chamber music from the same period. Even when relatively early works were published later,
like towards 1800, no dynamics were included.
Haydn - Duo for Two Violins in G, op 6 no 1 (vl primo) (n.d.)
Amsterdam: J.J. Hummel, n.d. (c.1796)
Clearly, composers were expecting their performers (often themselves, anyway) to deal with
various issues of expressivity, among which dynamics, independently. This situation did at
first not change with the advent of classicism, in which composers like Mozart or Clementi
were instrumental in creating a powerfully rhetorical, theatrical and with that, dynamical style.
Clementi Flute Trio op. 22 no 1 (a.k.a. op 23) (1789)
First ed. Berlin: J.J. Hummel [n.d., ca 1790]
Mozart's Sonata for four hands in C major, K 19d, written in England and very likely
performed there on harpsichords as well as on fortepianos, has no dynamical markings. This
piece, composed for his own use, is written down very efficiently as shown for instance by the
fact that the Alberti Bass in the rondo repeat is in short hand notation (from line 5, m. 7).
7
Mozart. Sonata four hands K. 19d, (1765). Third movement (primo)
First edition: De Roullede, n.d. (ca.1770).
This example notwithstanding, Mozart was one of the first composers to start notating
dynamics in almost every keyboard work. This doubtlessly had to do with the fact that his
works were written with publication in mind. But even so, some of his own (as well as of
contemporaries) later repertory for dynamical keyboard instruments still did not have dynamic
markings. The Sonata in C major, K. 330 (1781-3, published in 1784), has very few markings,
and his late Sonata K. 545 (1788) - including a highly expressive second movement - has not
a single dynamic marking. The genesis of this Sonata, described by Mozart himself in his
thematic catalogue as "for beginners," may have something to do with that: it was not
published in Mozart's lifetime and first appeared in print in 1805.
Mozart. Sonata in C major, K. 545 (1788). First movement, mm. 1-7.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
8
Mozart. Sonata in C major, K. 545 (1788). Second movement, mm. 1-6.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
Until the middle of the 18th century, music was rarely written for publication. Performances,
in particular of orchestral music, stayed under the control of the composer.3 Clive Brown,
referring to Reichardt’s discussion of dynamics, remarks: 'In practice, very few composers in
the 1770s or 1780s took the sort of care with their scores that Reichardt recommended, and
very few orchestras seem to have achieved his ideal.' Reichardt observed in 1776:
If the composer wants to have it absolutely presicely performed he will do
well to include all these different gradations exactly under the notes where
they should occur... Or he must come to an understanding with his
orchestra on particular occasions, namely that every whole or half bar
should become to a degree brighter or darker, or whatever may be agreed.
From time to time, however, precise specification will nevertheless be
necessary so that one player does not get louder more quickly than another.
The composer may, therefore, merely write pp. in the first bar, in the
second cresc. and in the ninth ff.: in any case he may also add the sign <
and the orchestra will make the second bar p, the third poco p, the fourth
rinf, the fifth poco f, the sixth piú f, the eigth f and the ninth ff.4
While Reichardt's suggestion for a finer notation of dynamics was never realized, it remains
interesting to notice that this discussion was already alive and well in the early classical
period.
2. Character
Often, if there are any dynamic markings in these early scores, they are found at the opening
of the piece. They seem to indicate a character rather than a local decibel level or expressivity.
This means that the opening dynamic is multi interpretable, just like the character itself. As in
a theater play, the 'musical personality' is identified; but the theatrical persona itself has
complete freedom to realize, within this personality, a flexible dynamical scheme in order to
realize the rhetorical content of the music. The only dynamics one initially finds are forte and
piano. In the third quarter of the twentieth century the literal interpretation of these general
markings lead to the so-called 'terrace dynamics'. But C.P.E. Bach explains why these two,
3
Clive Brown. Classical and Romantic Performance Practice, 1750-1900. Oxford: Oxford University Press,
1999. p. 61.
4
Johann Friedrich Reichardt. Über die pflichten des Ripien-Violinisten. Berlin and Leipzig, 1776. pp. 65-67.
Quoted in Brown 1999, p. 62.
9
seemingly extreme markings, are sufficient as they carry the possibility of a much broader
range of dynamics:5
Es ist nicht Wohl möglich, die Fälle zu bestimmen, wo forte oder piano statt
hat, weil auch die besten Regeln eben so Viel Ausnahmen leiden als sie
festsetzen; die besondere Wirckung dieses Schatten und Lichts hängt von
den Gedancken, von der Verbindung der Gedancken, und überhaupt von
dem Componisten ab, welcher eben so wohl mit Ursache das Forte da
anbringen kan, wo ein andermahl piano gewesen ist, und offt einen
Gedancken sammt seinen Con-und Dissonanzen einmahl forte und das andre
mahl piano bezeichnet. Deswegen pflegt man gerne die wiederhohlten
Gedancken, sie mögen in eben derjenigen Modulation oder in einer andern,
zumahl wenn sie mit verschiednen Harmonien begleitet werden, wiederum
erscheinen, durch forte und piano zu unterscheiden.
In the early string quartet in G, K. 80, the first movement Adagio is marked p by Mozart.
There is no other marking in the entire movement. The second movement Allegro has but one
single f at the beginning.
Mozart. String Quartet in G, K. 80 (1770). First mvt Adagio, Second mvt Allegro
Urtext der Neue Mozart Ausgabe (1966).
It is the same in the early Missa Brevis, K. 49 by Mozart.
5
Carl Philipp Eamuanel Bach. Versuch über die Wahre Art das Klavier zu spielen. Part I, Chapter III: Vortrag.
First ed. 1753. §29, p. 129.
10
Mozart. Missa Brevis, K.49 (1768). Kyrie.
Urtext der Neue Mozart Ausgabe (1968).
Dynamics may indicate a tutti/solo exchange, as in this early work for fortepiano by JC Bach.
J. C. Bach - Sonata in Bb major, op 5 no 1 (1765)
First publ. 1768; this ed. London: Welcker, n.d. (ca.1775)
Leopold Mozart (1719–1787) stresses the importance of basing an interpretation on character
in his Versuch einer gründlichen Violinschule (1755, publ. 1756). The interpretation of the
basic touch as well as articulation, tone colour, pedaling, tempo and rubato, direction, tension,
11
in fact every element of the performance leaves its mark on dynamics. With this, all dynamics
depend on this elementary and most important choice.6
§7. Bevor man zu spielen anfängt muß man das Stück wohl ansehen und
betrachten. Man muß den Charakter, das Tempo und die Bewegung, so das
Stuck erfördert, aussuchen, und sorgfältig nachsehen, ob nicht eine Passage
darinnen stecket, die oft beym ersten Ansehen nicht viel zu bedeuten hat,
wegen der besondern Art des Vortrags und des Ausdrucks aber eben nicht
leicht abzuspielen ist. Man muß sich endlich bey der Ausübung selbst alle
Mühe geben den Affect zu finden und richtig vorzutragen, den der
Componist hat anbringen wollen; und da oft das Traurige met den
Fröhlichen abwechselt: so muß man jedes nach seiner Art vorzutragen
beflissen seyn. Mit einem Worte, man muß alles so spielen, daß man selbst
davon gerühret wird.
In his Klavierschule (1789), Daniel Gottlieb Türk (1750–1813) confirms:7
§31. Jede einzelne Stelle zu bestimmen, welche etwas stärker oder
Schwächer, als die vorhergehenden und folgende, vorgetragen werden muß,
ist schlechterdings unmöglich; indeß kann man im Allgemeinen annehmen,
daß die lebhaftern Stellen eines Tonstückes stark, die zärtlichen singbaren
etc. aber schwächer gespielt werden, wenn auch im ersten Falle kein forte,
und im zweyten kein piano angemerkt ist.
Türk goes one step further by describing in a very general way which characters receive
which dynamics. Most interesting is a comment ('Anmerkung') underneath this paragraph
where he warns against an absolute interpretation of a written dynamic.8
§30. Ueber die jedesmal erförderliche Stärke des Tones begnüge ich mich
überhaupt anzumerken, daß die Tonstücke von einem muntern, freudigen,
lebhaften, erhabenen, prächtigen, stolzen, kühnen, muthigen, ernsthaften,
feurigen, wilden, wüthenden etc. Charakter alle einem gewissen Grad der
Stärke erfordern. Dieser Grad muß sogar noch vermehrt oder vermindert
werden, je nachdem die Empfindung oder Leidenschaft heftiger oder
gemaßigter dargestellt wird. Welche Grade der Stärke werden schon im
Ganzen hierzu erfordert. Und nun denke man sich, daß in jedem Tonstücke
selbst wieder verschiedene Anstufungen nöthig sind, die alle mit dem
Ganzen in einem angemessenen Verhältnis stehen müssen. Ein forte in
einem Allegro furioso muß daher ungleich stärker seyn, als in einem
Allegro, worin nur ein gemäßigter Grad der Freude herrscht u.s.w.
Die Tonstücke von einem sanften, unschuldigen, naiven, bittenden,
zärtlichen, rührenden, traurigen, wehmütigen etc. Charakter erfordern
insgesammt einen schwächern Vortrag. Der Grad der Stärke muß aber der
jedesmaligen Empfindung genau entsprechen, und folglich auch in den
meisten nur eben genannten Fällen verschieden seyn. So wie bey stark
vorzutragen Tonstücken, außer der ihnen überhaupt zukommenden Stärke,
immer noch ein höheres Grad zum fortissimo möglich seyn muß, eben so
muß auch bey den schwach vorzutragenden Tonstücken noch ein piano und
pianissimo angebracht werden können.
6
Leopold Mozart. Versuch einer gründlichen Violinschule. Salzburg: 1756/Dritte vermehrte Auflage. Augsburg,
1787. Ch. 12, §7.
7
Daniel Gottlieb Türk. Klavierschule. (1789). Ch. 6 Part III, §31. p. 349. Original quote see Appendix 1.3.
8
Türk Klavierschule. Part III, §30. p. 349.
12
Anm. 1. Oft bestimmt der Komponist den Hauptgrad der Stärke oder
Schwäche durch die zu Anfange beygefügten Worte sempre forte oder
sempre piano. Dieses sempre darf aber nicht in einer strengen Bedeutung
genommen werden; denn der Tonsetzer will dadurch nur so viel sagen, daß
der Vortrag im ganzen stark oder schwach seyn soll. Einzelne Gedanken
müssen dessen ungeachtet dem Affekte gemäß modificirt (stärker oder
schwächer vorgetragen) werden.
It follows that not just detailed dynamics, but also the basic dynamic level of a piece is
determined by its character.
Many scores before 1800 lack an opening dynamic. But although an opening allegro may not
have a dynamic marking in bar one, the performer is nevertheless supposed to start with the
dynamic related to that character - in other words: a dynamic which successfully brings out
the chosen character. In the language of the early classical style (ca 1750-1790) this would
mean that for a lively allegro the marking forte is implied (the interpretation of that marking,
or how loud one really has to play, is then determined from the character).
Typically, after a lacking opening dynamic for an allegro or presto the second motive has a
contrasting marking. The next motive (sometimes a return of the opening motive) may be
marked forte again.
C.P.E. Bach. Sonata in E minor, Kenner und Liebhaber V, Wq. 59/1 (1785). First movement.
Krebs: Breitkopf und Härtel, ca 1895.
The presto as well as passionato character of this sonata are sufficient reason for the
performer to choose the correct basic dynamic of forte (which is, again, not an absolute
dynamic level but depends on factors like the performer's personality, the instrument,
performance venue, size of the audience, etc).
C.P.E. Bach. Sonata in A minor, Kenner und Liebhaber III, Wq. 57/2 (1781). First movement.
Krebs: Breitkopf und Härtel, ca 1895.
Similarly, slow movements often have no piano marking.
Haydn. Sonata in D, Hob. XVI/37, second movement (1780).
Henle Urtext (1972)
13
3. What if there is an opening dynamic?
Mozart, who carefully prepared his sonatas for publication was one of the first to indicate the
hitherto lacking 'normal' opening dynamic, initially thought to be obvious enough to justify its
absence.
Mozart. Sonata in C, K. 279, first movement. 1775.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
However, also when the character of a movement is unusual the composer would have to
indicate the required dynamic in order to make his intentions clear. Therefore, when dynamics
appear at the opening of a movement they often do not coincide with the 'normal' dynamic of
the movement. Carl Philipp Emanuel Bach's Sonatas, Rondos and Fantasies für Kenner und
Liebhaber are a treasure-trove for refined and sometimes unexpected dynamics. In each case,
Bach made sure the performer has no doubts about his intentions, for instance when he asks
for a forte opening of a Larghetto or a piano opening of an Allegretto.
C.P.E. Bach. Sonata in G, Kenner und Liebhaber II, Wq. 56/2 (1780). Second movement.
Krebs: Breitkopf und Härtel, ca 1895.
C.P.E. Bach. Rondo in C, Kenner und Liebhaber II ,Wq. 56/1 (1780).
Krebs: Breitkopf und Härtel, ca 1895.
Mozart loved surprising his audience this way, as shown by his letter on the Paris Symphony
K. 297, in particular about the dynamical effect in his last movement Allegro: 9
Having observed that all last as well as first allegros here begin together
with all the other instruments, and generally unisono, mine commenced
with only two violins, piano for the first eight bars, followed instantly by a
forte; the audience, as I expected, called out "hush!" at the soft beginning,
and the instant the forte was heard began to clap their hands.
9
Mozart. Letter to his father from Paris. July 3, 1778. Gutenberg eBook. The Letters of Wolfgang Amadeus
Mozart, Vol. 1./ Transl. Lady Wallace. 2004.
14
A similar effect is found in the last movement of the A minor Sonata K. 310 (also written in
Paris in 1778), where the forte is postponed and the tension is reinforced by the piano opening.
Mozart. Sonata in A minor, K. 330, second movement. 1775.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
But even from his earliest sonatas both Mozart's fast and slow movements showed unusual
dynamics, as in the andante of his first published Sonata K. 279 which starts forte.
Mozart. Sonata in C, K. 279, second movement. 1778.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
Starting a fast movement, p could indicate a special effect. Mozart used in several of his
sonatas the contrast between the imaginary soloist and the orchestral tutti, like in his first
Sonata, K. 279 (last movement) or in last movement of the Sonata in Bb major K. 333, where
the 'orchestra' enters in m. 8.
15
Mozart. Sonata in Bb, K. 333 (1783). Third movement, mm. 1-16.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
As mentioned above, there are no dynamics in Haydns keyboard sonatas until the six
Auenbrugger Sonatas (1780). In these sonatas the dynamics seem to indicate the passages the
character of which otherwise might be misunderstood.
4. Interpreting lacking dynamics
Acknowledging that opening dynamics are never really lacking (because they are implied)
can lead to interesting discussions, as in the cases where a composer brings the opening theme
or motive back - but not in the supposed opening dynamic. In this case the performer will
have to apply reason in order to be able to decide about the actual opening dynamic.
C.P.E. Bach. Rondo in D, Kenner und Liebhaber II, Wq. 56/3. (1780).
Krebs: Breitkopf und Härtel, ca 1895.
In the Rondo in D major, Wq. 56/3, the opening motive returns in m. 5 as a p motive but at
the beginning of the second phrase in f (assuming that the f at the phrase climax in m 7 applies
16
to the opening of the second phrase in m. 9, which is the opening motive, ornamented and
therefore more emphatic). Which of these dynamics applies to the opening motive? As the
second motive of the piece (upbeat to m. 2) is marked p, it seems logical - in absence of other
dynamics than f and p until much further in the piece, when we see a pp marking - to play the
opening in the contrasting character of forte. In this scenario, the halfway point of the phrase
(upbeat to m. 5) contrasts with the opening and is therefore marked p, no matter that it is the
same motive. After all, CPE Bach himself mentioned how the same motive can be heard in
contrasting dynamics: "...die besondere Wirckung dieses Schatten und Lichts hängt von den
Gedancken, von der Verbindung der Gedancken, und überhaupt von dem Componisten ab,
welcher eben so wohl mit Ursache das Forte da anbringen kan, wo ein andermahl piano
gewesen ist."10
Mozart's interesting indication at the opening of the slow movement of the C minor Sonata
leaves room for speculation, as to the difference in nature between p and sotto voce. The
opening is doubtlessly meant to be played p; but the sotto voce independently refers to the
character - it adds a level of intimacy which has implications for instance for the balance
between the left and the right hand. This is interesting as Mozart does not rely completely on
the character indication of the dynamic indication p anymore: he adds a new layer of meaning
to the description of the opening theme.
Mozart. Sonata in c minor, K. 457 (1784). Second movement.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
10
See footnote 4.
17
II. Understanding Dynamics
1. The performer's domain
Treatises of the second half of the 18th century treat dynamics as the domain of the performer.
Carl Philipp Emanuel Bach summarizes the elements of performance, and with that the
performer's responsibilities:
Die Gegenstände des Vortrages sind die Stärke und Schwäche der Töne,
ihr Druck, Schnellen, Ziehen, Stoßen, Beben, Brechen, Halten, Schleppen,
und Fortgehen. Wer diese Dinge entweder gar nicht oder zur unrechten
Zeit gebrauchet, der hat einen schlechter Vortrag.11
Dynamics are mentioned often as an indispensible expressive tool, to be understood in the
context of the musical language. In the words of Leopold Mozart:
From this it follows that the prescribed piano and forte must be observed
most exactly, and that one must not go on playing always in one tone like a
hurdy-gurdy. Yea, one must know how to change from piano to forte
without directions and of one's own accord, each at the right time; for this
means, in the well-know phraseology of the painters, Light and Shade.12
Johann Joachim Quantz says that one of the duties of the accompanist is to understand the
dynamics. He recommends the fortepiano over the harpsichord, since it is so much easier to
follow the dynamics of the soloist (the exact use of which the keyboardist has to decide about
himself):13
[...] on a pianoforte everything required may be accomplished with the
greatest convenience, for this instrument, of all those that are designated by
the word keybord, has the greatest number of qualities necessary for good
accompaniment, and depends for its effect only upon the player and his
judgement.
At the same time, be careful to not do too much:
The forte and piano must never be unduly exaggerated. [...] You must
always be able, in case of necessity, to express an additional Fortissimo or
Pianissimo. It may often happen that you must unexpectedly bring out or
soften a note, even if nothing is indicated. [last italics mine]
Johann Adam Hiller in his Anweisung zum musikalisch-zierlichen Gesänge (1750) explains
that the dynamics are fully the domain of the singer:14
§8. Aus dem, was oben über das Tragen der Stimme gesagt worden, ergiebt
sich, daß ein Ton, der von einiger Dauer ist, immer am Ende etwas stärker
gehört werden müsse, als im Anfange. Daß dieser Unterschied so groß
11
Bach Versuch. Part I, Chapter III: Performance. 1753/2nd ed 1787. p. 87.
Leopold Mozart Violinschule. Ch. 12, §8.
13
Johann Joachim Quantz. On Playing the Flute. Translation of Versuch einer Anweisung die Flöte Traversiere
zu spielen. Berlin, 1752/Translated by Edward R. Reilly. New York, Schirmer, 1985. Chapter 17: Von dem
Pflichten derer, welcher accompagniren, p. 175.
14
Johann Adam Hiller. Anweisung zum musikalisch-zierlichen Gesänge (Leipzig, 1750). Ch II, §8, p. 22.
12
18
seyn solle, als er zwischen piano und forte zu seyn pflegt, ist damit nicht
gesagt; es giebt, zwischen beyden, so viele Mittelgrade, daß wir bey weiten
nicht Namen genug haben, sie anzudeuten. Diese alle muß ein guter Sänger
in seiner Gewalt haben [...]
Indeed, his own songs have no dynamical indications whatsoever.15
Hiller. Der Schöne Apfel.
From: Neefe. Vademecum fur Liebhaber des Gesangs und Klaviers. Leipzig, 1780.
His larger works, like the "Kantata auf die Ankunfft der hohen Landesherrschaft" (1765) have
only p and f markings; sometimes close together. From the quote of Hiller's treatise (above) it
appears that these are general markings to be modified according to circumstances,
understanding, taste etc: "there are so many intermediary dynamical grades that we do not
have enough names for all of them".
Hiller. Kantata auf die Ankunfft der hohen Landesherrschaft (1765).
Türk leaves no doubt as to the responsibility of the performer in the area of dynamics:16
15
Johann Adam Hiller. Der Schöne Apfel. Uit: Neefe. Vademecum fur Liebhaber des Gesangs und Klaviers.
Leipzig, 1780.
16
Türk Klavierschule. Ch. 6 Part III, §29. p. 348.
19
Auch bey der sorgfältigsten Bezeichnung ist es nicht möglich, jeden Grad der
erforderlichen Stärke und Schwäche zu bestimmen. So viel wir auch Worte dazu
haben, so sind sie doch bey weiten noch nicht hinreichend, alle mögliche
Abstufungen anzuzeigen. Der Spieler muß daher selbst fühlen und beurtheilen
lernen, welchen Grad der Stärke and Schwäche der jedesmal auszudruckene
Charakter erfordert. Das beygefügte forte und piano bestimmt den Ausdruck nur
so ungefähr und im Ganzen; wie überhaupt würden aber diese Worte beygefügt
werden müssen, wenn jede einzelne Note, welche eine besondere Schattirung
verlangt, damit bezeichnet werden sollte.
And even at end of the 18th century, Johann Peter Milchmeyer (1750-1813) mentions in his
treatise Die Wahre Art das Pianoforte zu spielen (1797) that the notation of dynamics is
incomplete and can only be understood fully by a 'very accomplished performer'.17
Jeder Ausdruck der Stärke oder Schwäche hat wieder seine verschiedenen
Grade, welche man in der Musik selten bezeichnet, und die nur ein sehr
geschickter Spieler ausdrücken kann.
2. Understanding the musical language
While in the 18th century dynamics are evidently considered to be part of the expressive
domain of the performer, there are many ways in which composers did indicate dynamics
implicitly. Many of these implications are straightforward and in fact described in most
treatises, even though modern day performers may not always take the hint. The crucial factor
is that understanding the implications for dynamics in the score follows from the way we
understand the musical language. In short: correct understanding of the musical language
makes it possible for the performer to understand how to use and change dynamics without
receiving a specific indication.
Shaping the musical language by way of dynamics is done on three levels.
a. Diction or articulation: shaping the musical word (motive).
b. Phrasing: shaping the musical phrase.
c. Structure: shaping the musical story.
a. Diction
In order to 'speak' each musical 'word' (the musical motive) correctly we need to design the
correct emphasis (both accentuation and length). This accentuation depends on the relative18
harmonic tension within a musical motive, the gesture that governs that motive, the
articulation defining the diction of the gesture, and the dissonance and consonance within that
musical word as related to the local harmony. Furthermore, the motive is placed strategically
within the bar which means that rhythm and meter play an important role, as they determine
the relative importance, i.e. the accentuation of the beats.
ADD EXAMPLE FROM HAYDN SONATA C HOB XVI/50, DEVELOPMENT
17
Johann Peter Milchmeyer. Die Wahre Art das Pianoforte zu spielen. Dresden, 1797. Ch. IV: Vom
Musikalischen Ausdruck, p. 48.
18
'Relative' tension because in a situation with shifting harmonic center harmonies will acquire new meaning and
develop new relationships to each other. Harmonies which were consonant may become dissonant; and that
meaning can only be determined by their interaction.
20
Generally speaking, the first beat has the greatest weight. It is on first beats that one usually
finds the most important dissonances. Release of that dissonance generally takes place on an
off-beat. In case a motive lasts for more than one bar and therefore contains two downbeats,
the performer must find out which of the two beats is the 'good' and which one is the 'bad'
downbeat.19
There is no classical repertory in which the bars are not paired in either a good and bad, or in
a bad and good bar; or in four bar groups in which the third is the heavy bar. The standard
classical phrase has a four bar second half (2+2+4). In addition, four bar groups occur often in
development sections where they create a larger flow.
Exceptions are irregular structures in which motives have for instance three bars, leading to
phrases of for instance 7 or 9 or another, often uneven number of bars.
b. Shaping the phrase
Any musical phrase has but one climax. In order to understand where this climax takes place
the performer needs to consider first of all the harmony, next the harmonic rhythm, and finally
the melody.
The standard classical phrase has a regular structure of 8 bars (2+2+4), in which there is a first
small highpoint in m. 3 or on the downbeat of m. 4 with the phrase climax in m. 7. While
towards 1800 this climax often is identified by the composer with dynamic markings and/or
accentuation, there is no such road map in the earlier classical works. The working of the
harmony being of prime importance, the performer would understand that the grammar of the
phrase necessitates a dynamic development, resulting in a crescendo-diminuendo towards and
from the highpoint.
Haydn. Sonata in C minor, Hob XVI/20 (1771). First movement, mm. 1-11.
Henle Urtext (1972).
19
'Gut' and 'Schlicht', translated erroneously by 'good' and 'bad' instead of 'good' and 'slight'. In everyday practice
the terminology has been standardized to 'good' and 'bad'.
21
The first motive - 2 bars - of the opening theme of Haydn's C minor Sonata Hob XVI/20
consists of four slurred gestures. The first starts on a consonance on the tonic; the second, one
step higher on the scale, is more dissonant and therefore receives more stress: the first bar has
a crescendo, although each slur in itself is diminuendo.20 The third gesture starts on the
downbeat of m. 2 and is the heaviest of the four, with the most expressive and relatively
loudest start of the slur and the largest diminuendo towards the end of the slur. Bar 2 is
therefore the good bar; bar one the bad bar. The fourth gestural slur, although on the less
heavy third beat of the second bar, nevertheless starts with the most important and most
dissonant harmony of this two bar motive: I6/4, the arrival point, but not quite as loud as the
downbeat of the second bar. Its release and diminuendo to V7 on the forth beat completes the
first two bars of the opening phrase: the second bar has an overall diminuendo.
The next two bars build up in a similar way through bar 3 and arrive at the first, lesser
expressive climax of the 8 bar phrase on the downbeat of bar 4, illustrated by the expressive
notation of the dissonance on the downbeat (the appogiatura), aided by the Praller on the
second eight note of the bar which lifts up the descending gesture and reinforces the effect of
release. Bar 5 seems to parallel bar 1, but the ascend of bar five continues in bar 6 and
develops into a rhetorical question to end on the prototypical (because a rhetorical question)
I6 on the third beat. All three ascending, questioning figures may be subsequently louder.
Together with the rocket upbeat (with great direction but itself diminuendo) this is the perfect
rhetorical set-up for the climax of this phrase on the downbeat of bar 7 (the third bar of the 4
bar group 5-6-7-8), after which each following gesture works towards a release of the tension:
bar 7-8 is diminuendo.
c. Shaping the story
The structure of the piece equals the 'way in which the story unfolds': how the phrases follow
each other, the development of the harmonic tension from one phrase to the next, how the
various characters interact (the rhetorical 'dialogue'). As a result the tension - and with that,
the dynamics - rises and decreases and may eventually flow away at points of arrival.
20
See Ch. III/3, The correct performance of the slur.
22
III. Implied dynamics.
1. Introduction
In any score for any instrument, dynamics are implied in many ways. As in the harpsichord
style, in early piano repertory the number of notes has an impact on the volume: more notes
means louder, fewer notes means softer. Similarly, deeper (as well as more) basses allow for
greater volume, while melodies in octaves have a potential to be louder; reversely, it is safe to
assume that a classical composer asks for a louder context when the basses are deeper and the
melody is written in octaves.
Quantz mentioned these techniques with regards to the harpsichord in his chapter on
accompaniment:21
On a harpsichord with one keyboard, passages marked Piano may be
produced by a moderate touch and by diminishing the number of parts,
those marked Mezzo Forte by doubling the bass in octaves, those marked
Forte in the same manner and also by taking some consonances belonging
to the chord into the left hand, and those marked Fortissimo by quick
arpeggiations of the chords from below upwards, by the same doubling of
octaves and the consonances in the left hand [...]
In an earlier paragraph - the first one in which Quantz mentions the pianoforte, and with that
makes his the first treatise to do so - Quantz stresses once more how the application of
dynamics is the domain of the performer.22
Das stark und schwach Spielen kann zwar auf dem Clavicymbal oder
Flügel, besonders wenn derselbe nur ein Klavier hat, nicht so ab- und
zunehmend ausgedrücket werden, als auf dem Instrumente, welches man
Pianoforte nennet, allwo die Saiten nicht mit Federn gerissen, sondern
durch Hämmer angeschlagen werden: dessen ungeachtet aber, kömmt doch
bey den Flügel, viel auf die Art des Spielens an.
This style of playing and creating dynamics carried over from the harpsichord to the early
fortepiano style. Only after 1800 these harpsichord techniques vanished little by little from
piano playing.
Ornamentation in the form of trills or appoggiaturas which stress a dissonance or de-stress a
consonance on a harmonic release are important in the early classical style, as shown in every
one of the works of CPE Bach. These ornaments still play a role in the works of Haydn, and
increasingly less in the works of Mozart and Beethoven.
21
Johann Joachim Quantz. On Playing the Flute. Translation of Versuch einer Anweisung die Flöte Traversiere
zu spielen. Berlin, 1752/Translated by Edward R. Reilly. New York, Schirmer, 1985. Section VI Of the
keyboardist in particular, §17, p. 259.
22
Quantz Versuch. Section VI Von dem Clavieristen insbesondere, §17, p. 259.
23
C.P.E. Bachs Empfindungen. Freie Fantasie für Klavier, H.300, Wq.67 (1787).
Mainz: Kreuz, 1950.
In the 'Empfindungen', perhaps CPE Bach's most important work for solo keyboard, the
ornaments must be understood in the realm of dynamics. The turn on the first D deemphasizes a release from the dissonant Eb to D. The turn on the second D stresses the
increasing dissonance of the repeated D on the I6/4 harmony, becoming V7 on the second line
and finally releasing to G minor (second line, beat 2).
How dynamics may be implied by texture is still apparent in Beethoven's Sonata op 10/3 in D
major. On the first page of the first movement Beethoven writes but one crescendo and not a
single diminuendo. Nevertheless, if texture and articulation are taken into account, a pattern
of flexible dynamics arises from the page.
Beethoven. Sonata in D, op 10/3, first movement (1796/98).
First edition: Eder, n.d.
The opening phrase has no crescendo but in bar 2 the double unisons become quadruple
unisons, not just implying but positively causing a crescendo. The sf on the climax in m. 4 is
contextual, meaning that the actual volume is decided by the performer and depends on
instrument, tempo, drive, and the desire to stay largely within the p character or not; but
doubtlessly this sf is the summit of a (perhaps minimal) crescendo.
24
The second phrase is legato. Its climax is on the dissonant d# syncopation in m. 8 (the last
note of line 1). It is separated from the notes before, meaning that the articulation gives extra
life to the dissonance, and its tension releases in the next bar. The tension build-up towards
the d# implies a crescendo, again within the p character. But the build-up of tension must be
convincing enough to allow for the third phrase to be forte. In the second half of this forte
phrase another crescendo is implied by the fact that the hands of the pianist diverge (a
rethorical gesture often found in the classical style, implying a growing: the right hand rising
and the left hand descending) and create a crescendo, increasing the potential for forte. The
climax of this phrase is in m. 15 (line 2, m. 7). The D chord in m. 16 is a return to the tonic
and therefore has less tension than the dominant on m. 15. This means that m. 16 is a
(relative) release and therefore is slightly less loud.23
The fourth phrase however, has a notated crescendo. Now the performer is asked to leave the
p character, aided again by the increasing number of unisons, and create three consecutive ff
accents, the last one forming a climax on the quadruple24 f# in m. 22 (3rd line, m. 6).
2. Dissonance and consonance
Difference between a dissonance and a consonance is the most basic natural cause of
changing dynamics. Before 1800, this is never indicated as it reflects the normal accentuation
of speech. Every treatise before 1800 mentions that a dissonance is to be played louder than
the resulting consonance:
The more, then, that a dissonance is distinguished and set off from the
other notes in playing, the more it affects the ear. But the more displeasing
the disturbance of our pleasure, the more agreeable the ensuing pleasure
seems to us. Thus the harsher the dissonance, the more pleasing is its
resolution.25 Some of the dissonances have more effect, and so must be
struck with greater force than the others.26
It is important to understand the four basic positions of the dissonances and consonances on
the first two levels of musical language, diction and phrasing:
1. A melodic consonance on top of a harmonic consonance - very little tension on both levels
3. A melodic dissonance on top of a harmonic consonance - rather little tension, the melody
gains expressivity
2. A melodic consonance on top of a harmonic dissonance - relatively high tension in the
context; the melody may be full of tension or may within the context be rather dolce
23
The bass note EE is the climax of the phrase and is doubtlessly meant to be an octave like the basses before.
Like the highest f# (missing in this edition in m. 22), which was at the time the second highest note of
Beethoven's piano, this EE was the lowest. As many pianists did not yet own a piano with this extended compass,
the first edition does not print these notes. Correctly, Henle gives the option to play the lowest EE as well as the
highest f# (and prints them in brackets).
24
The prudent publisher of this sonata stayed within the five octave range of the standard instrument of his time.
However, new instruments had already been equipped with the f#''' en g''' and therefore Beethoven doubtlessly
implied the continuation of the octaves.
25
Johann Joachim Quantz. On Playing the Flute. Translation of Versuch einer Anweisung die Flöte Traversiere
zu spielen. Berlin, 1752/Translated by Edward R. Reilly. New York, Schirmer, 1985. Ch. XVII section VI: 'Of
the Duties of those who accompany a concertante part'. §12, p. 227.
26
Quantz. On Playing the Flute. Ch. XVII section VI: 'Of the Duties of those who accompany a concertante part'.
§13, p. 227.
25
4. A melodic dissonance on top of a harmonic dissonance - high harmonic and melodic
tension as well as expressivity
Each of these four position implies a certain dynamic level which depends on the way the
performer interprets the context. This is not to say that a performer is at liberty to interpret, or
change the meaning of a harmony - because the harmony, the grammar of the piece, cannot be
treated in other way than what it is - but the relative tension (relative to the harmony, that is)
of the melody leaves room for interpretation. This is why before 1800 most of these dynamic
fluctuations are not indicated: it is the domain of the performer.
Carl Philipp Emanuel Bach mentions (like in fact every treatise before 1800) how dissonances
and consonances each have their own dynamics27:
Indessen kan man merken, daß die Dissonanzen insgemein stärcker und die
Consonanzen schwächer gespielt werden, weil jene die Leidenschaften mit
Nachdruck erheben un diese solche beruhigen, Fig. XIV (a). Ein
besonderer Schwung der Gedancken, welcher einen heftigen Affect
erregen soll, muß starck ausgedruckt werden, gemeiniglich forte (b). Man
kan allenfalls auch diese Regel mercken, welche nicht ohne Grund ist, daß
die Töne eines Gesangs, welche ausser der leiter ihrer Ton-Art sind, gerne
das forte vertragen ohne Absicht, ob es Con- oder Dissonanzen sind, und
daß gegentheils die Töne, welche in der Leiter ihrer modulirenden Ton-Art
stehen, gerne piano gespielt werden, sie mögen consoniren oder dissoniren
(c)
C.P.E. Bach Versuch (1753)
Every work by Carl Philipp and his contemporaries must be interpreted with this fluctuation
in tension - and therefore, in dynamics - in mind. The Sonata in G major from the fourth
volume of Kenner und Liebhaber opens on the first downbeat with a melodic dissonance on
top of the harmonic dissonance (V); the second downbeat has a melodic dissonance on top of
a harmonic consonance (I). The downbeat of bar 2, the 'bad' bar, is therefore more relaxed and
as a result less loud than the downbeat of bar 1, the 'good' bar.
C.P.E. Bach. Sonata in G, Kenner und Liebhaber IV, Wq. 58/2. (1783).
Krebs: Breitkopf und Härtel, ca 1895.
27
Bach Versuch. Part I, Chapter III: Vom Vortrage. §29, p. 129.
26
Türk finally reminds us where the dissonances are to be found. They include a) notes which
are themselves dissonances with the bass; b) those notes which prepare (through a slur) a
dissonant interval; c) syncopated notes; d) notes which do not belong to the key in which one
plays or has modulated into; e) notes which stand out because of their length or pitch; notes
which become important because of the harmony underneath, etc.28
och gibt es verschiedene einzelne Töne, welche mit Nachdruck vorgetragen
werden müssen. Hierunter gehören, außer den Vorschlägen, vorzüglich
diejenigen Intervalle, die sich zu dem Basse etc. selbst wie Dissonanzen
verhalten a), oder durch welche (vermittelst einer Bindung) dissonirende
Intervalle vorbereitet werden b); ferner die synkopirten Noten c), die
Intervalle, welche nicht zur diatonischen Tonleiter desjenigen Tones
gehören, worin man modulirt d), die Töne, die sich durch ihre Länge, Höhe
oder Tiefe etc. merklich auszeichnen e), die Intervalle, welche durch die
zum Grunde liegende Harmonie wichtig werden f) u.s.w.
Türk Klavierschule (1789)
Türk adds to the discussion that often there is no slur between a dissonant-consonant release,
as it is simply not necessary to indicate the obvious: the performer knows what to do, as in the
diction of speech.
3. The correct performance of the slur
Mozart and Beethoven knew their pianos intimately. The refinement as well as the vigour and
drama which must have characterized their performances, the 'sound effects' they were able to
realize, as well as aspects of their technique which were eminently suited to their own pianos
such as pedaling, articulation, and fingering, and certainly also dynamics can be understood
and retrieved to some extent when their music is played on the pianos of their time.
The rapid decay of the fortepiano tone, and the resulting relative lack of singing power, is
perhaps the most audible difference between the fortepiano and the modern piano. On a
modern piano the tone has a soft attack, develops slowly, and sings for a long time before
gradually dying away. The fortepiano tone has a sharp attack and fades quickly, which makes
a natural articulation with the next tone possible and logical. The fortepiano touch is therefore
28
Türk Klavierschule. Ch. 6 Part II, §15. p. 337.
27
fundamentally non-legato. If a performer on a five octave fortepiano (all repertory from the
period ca 1750-1800) wants to play a sounding legato, the only way to do it is by playing
diminuendo: the new sound of each next note then falls into the rapidly decaying sound of the
previous note.
In his Versuch einer gründlichen Violinschule of 1755 (published in 1756) Leopold Mozart
makes an enormous effort to impress on the reader the correct execution of the slur. The
inherent dynamical shape (as well as its implications for rubato, which is the subject of
another study29) of the slur is so important to Mozart that he explains it in various words no
less than 11 times within the treatise. Here are four of those passages:
The first of the two notes coming together in one stroke is accented more
strongly and held slightly longer, while the second is slurred on to it quietly
and rather late.30
[...] Now if in a musical composition two, three, four, and even more notes
be bound together by the half circle, so that one recognizes therefrom that
the composer wishes the notes to not to be separated but played singingly
in one slur, the first of such united notes must be somewhat more strongly
stressed, but the remainder slurred onto it quite smoothly and more and
more quietly. [....] It will be seen that the stress fall now on the first, now
on the second, or third crotchet, yea, frequently even on the second half of
the first, second, or third crotchet. Now this changes indisputably the whole
style of performance...31
The first of two, three, four, or even more notes, slurred together, must at
all times be stressed more strongly and sustained a little longer; but those
following must diminish in tone and be slurred on somewhat later. But this
must be carried out with such good judgement that the bar-length is not
altered in the smallest degree. The slight sustaining of the first note must
not only be made agreeable to the ear by a nice apportioning of the slightly
hurried notes slurred onto it, but must even be made truly pleasant to the
listener.32
Now if several notes of this kind follow each other, over which, two by two,
a slur be placed, then the accent falls on the first of the two, and it is not
only played somewhat louder, but it is also sustained somewhat longer;
while the second is slurred onto it quite smoothly and quietly, and
somewhat late. [...]
But often three, four and even more notes are bound together by such a slur
and half-circle. In such a case the first thereof must be somewhat more
strongly accented and sustained longer; the others, on the contrary, being
slurred on to it in the same stroke with a dimishing of the tone, even more
and more quietly and without the slightest accent.33
29
Rubato, among other way implied by slurs and described by Leopold Mozart, Türk, CPE Bach and many other
treatises, was the subject of the Royal Conservatory fortepiano class research during the 2013-2014 season.
30
Leopold Mozart. A Treatise on the Fundamental Principles of Violin Playing /Translation of Versuch einer
gründlichen Violinschule. Salzburg, 1756/Translated by Editha Knocker 1948/reprint of 2nd ed. Early Music
Series no. 6. Oxford: Oxford University Press, 1986. Ch. 7, section 1, §3.
31
Mozart Treatise. Ch. 7 section 1 §20. The slur is novel enough for his readers that it even needs explanation as
a sign.
32
Mozart Treatise. Ch. 7, section 2, §5.
33
Mozart Treatise. Ch. 12, §10.
28
It goes without saying that observing these instructions about the correct performance of the
slur will alter every performance as compared to a performance in which the performer feels
at liberty to choose the dynamic shape of the notes under a slur. Not only will the
performance following Mozart's rules generally be more articulated, clear and lighter of tone;
but the flow of the piece will change, the gesturing will be more outspoken and in particular
also the local expressive rubato (the beginning of each slure) will take the place of a broader,
more general (dynamic) expressivity. The result is a colourful classicism rather than a lush
romanticism.34
The most occurring slur is the two-note slur, the so-called seufzer (sospiro, sigh). The seufzer
became the key signifier to the entire classical musical language and plays a prominent role in
the music of Haydn, Mozart, Beethoven and all their contemporaries. But even after 1800 the
seufzer, although less in the foreground, can be found in music of all composers.
Leopold Mozart's two elements defining any slur are also crucial to the effect of the seufzer.
First of all, the first note (often dissonant) is louder than the second. The amount of
diminuendo determines the expressive content of each individual slur. Secondly, the first note
is longer than the second, which is (in Mozart's words) "slurred onto it quite smoothly and
quietly, and somewhat late". This leads to a refined rubato, which in a string of seufzers may
cause an lilting effect, somewhat like the French inégale. The sweetness and delicateness of
this figure explains why seufzers appear so abundantly in the clavichord works of CPE Bach
and early Haydn.
C.P.E. Bach's style depends for its expressivity to a large extend on the working of the slurs.
The expressive effect of the seufzer on the downbeat of the first bar of his Sonata in B minor
is reinforced by the de-emphazising ornament on the second note. Measures 5 and 10 show a
series of highly expressive, pleading sighs.
C.P.E. Bach. Sonata in B minor, Kenner und Liebhaber I, Wq. 55/3 (1780). First movement, mm. 1-10.
Krebs: Breitkopf und Härtel, ca 1895.
Haydn's early style is in some respects quite close to C.P.E. Bach's. Of the works in which
seufzers are important we may surmise that they are perhaps best performed on a clavichord.
The early Sonata in Bb major, Hob XVI/2 is, like all of the early Divertimentos and Partitas,
written in the clavichord style. Each bar with a series of four seufzers releases onto a bar with
one longer, more important and ascending seufzer (mm. 31-35). In m. 45-46 the
appoggiaturas (to be performed on the beat in order to bring out their dissonance) function in
the same way.
34
In long slurs a messa di voce may be applied, see my Ch. IV, §4.
29
Haydn. Sonata in Bb, Hob XVI/2 (1750s). First movement, mm. 31-46.
Henle Urtext (1971).
The second movement has even more expressive, ascending seufzers resolving to a tragic
descending one in the next bar, which in turn develops through a series of new seufzers
towards the most expressive figure, an ornamented seufzer on the second beat of m. 4. In the
process, the development of the series of seufzers is the key factor in deciding about the
dynamics.
Haydn. Sonata in Bb, Hob XVI/2 (1750s). Second movement, mm. 1-7.
Henle Urtext (1971).
Also Mozart, Beethoven, Schubert, Mendelssohn, Chopin, Schumann, Brahms, and their
contemporaries used the seufzer for the most expressive and intimate moments in their music.
Mozart's first published sonata K. 279 in C leans heavily on this expressive figure. For
instance, every bar from m. 18 to 25 consists of seufzers. Interestingly, Mozart has given
detailed dynamics for each, opposing those that create dissonance marked forte (like m. 18,
first beat) to those that resolve to a consonance marked piano (m. 18, second beat, and so on).
30
Mozart. Sonata in C, K. 279 (1775). Second movement, mm. 14-28.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
4. Register
The compass of the five octave fortepiano reflects, with a little give and take, the compass of
the human voice. The essentially rhetorical quality of the early classical style, namely to
imitate speech and with that to create a musical conversation, depends on the same qualities
as the human voice itself. Goethe wrote in 1790 about Haydn's string quartets: "One hears
four intelligent people in conversation and has the impression that one has gained something
from the discourse."35
Like the human voice, the early fortepiano has registers; the soprano is delicate, bright, and
transparant; the tenor (in which most of the melodies takes place) has a richer tone, the alto
and bass registers are increasingly rich and gain in volume as well as length of the tone. The
treble, however, does not have a long lasting tone. In addition, the delicacy of its timbre does
not accommodate loud playing by itself. Loudness in the high treble is only possible if the
texture allows for it, e.g. a loud and full bass, or a moving and active middle voice, a
harmonic pattern which allows for ample use of the pedal, and/or a melody in octaves.
Examples are found on virtually every page of the keyboard works of CPE Bach, Haydn,
Mozart and their contemporaries. In each case decisions about the dynamics must be made
with care when performed on a contemporary fortepiano, keeping the nature of the higher
registers in mind.
Carl Philipp Emmanuel Bach Fantasies abound in passages without barlines. The barlines
lacking does not mean that these passages are entirely free since rhythm, accentuation, diction
and even dynamics - all based on the underlying harmony - are determining structural factors.
The Fantasy in A major from the fourth volume of the Kenner und Liebhaber series opens
with such a passage.
35
Goethe in a letter to Carl Friedrich Zelter, 9 Nov. 1829. Thanks to Professor James Webster (Cornell
University, Ithaca, NY) for this reference.
31
The way the notes are grouped shows how Bach wants the emphasis in this florish to be
applied. The performer must decide whether the first note of every group (with some
exceptions 8 notes per group) is a dissonance or a consonance against the prevailing harmony.
In line 1, for instance, the third, fourth and seventh group start with an appoggiatura; the first,
fifth and sixth with a consonant against A major. The second group only is ambiguous and
may be treated as a 'bad' beat. It is this dissonant/consonant opposition which gives meaning
by way of life, rhythm, and expressivity to this passage.
In the entire florish not a single high note asks to be emphasized: all of the high notes come
right after a 'beat' or in the middle of a group. This is not to say that they can't be expressive:
the expressivity, however, is to be dealt with in a delicate tone, rather than a romantic, full and
loud tone. For instance, a performance in which the e'' (line 1, group 3), a'' (line 3, group 1) or
d''' (line 4, group 2) are brought out forte after a preceeding crescendo would entirely be at
odds with early classical aesthetics. The d''' is particularly light: it is a release after arriving on
the low bass B.
CPE Bach. Fantasy in A, Kenner und Liebhaber IV, Wq. 58/6 (1783).
Krebs: Breitkopf und Härtel, ca 1895.
Haydn's retorical Sonata in Ab major, Hob XVI/46 (1768) is a prime example of classical
figuration. The bare first note Ab is ornamented to receive less of an attack; the high note Eb
on beat three is ornamented from above to soften the interval. The trills in bar two and three
de-emphasize the release of the three seufzer gestures. The sixteenth note gesture in bar 4
starts with a mild leap from the first to the second note. This is typical for the period: the high
note is lifted rather than attacked since the heavy downbeat is right before. The same lift of
high notes as in m. 1 appears in measures 5, 6 (the highest notes are also the lightest notes)
32
and 7. Measure 9 seems to start on a high note, but if the appoggiatura is performed as
described by every single treatise of the 18th century, namely on the beat and therefore
heavily, the high c''' is lifted and can be performed in accordance with the character of the
second phrase, namely delicato. In the repeat (m. 11ff) where the appoggiatura is replaced by
an ascending scale upbeat which receives a diminuendo. Therefore also this high c''' downbeat
is light and delicato.
Haydn. Sonata in Ab, Hob XVI/46, first movement (1768).
Henle Urtext (1971).
The end of the page is an extended dominant and therefore must be stressed with an
increasing tension. This is more easily achieved in the lower register than in the high register
and Haydn has therefore choosen this darker and more powerful passage to take place below
middle c.
As in all of his keyboard works, most high notes on the first page of Mozart's Sonata in Bb
major, K. 333 do not fall on a heavy beat. Typically, a high note falls immediately after a beat
and is therefore light and lifted; or before a beat and therefore functions as an upbeat, which
makes it light as well (mm. 1, 2, 4, as well as 13, 14, 15 etc).
A special case is when a note is repeated, usually immediately before and after a barline. the
note before the barline is consonant as well as light (end of the bar), the note after the barline
is dissonant and extra heavy because of the downbeat (mm. 5, 7, 17).
33
Mozart. Sonata in Bb, K. 333, first movement (1783).
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
Of course, Mozart knows how to create a climax on a high note. However, this climax has to
meet certain requirements. First of all it will need to take place on a heavy beat. Secondly, this
beat must be sufficiently supported to allow for the high note to sound climactic, i.e. loud and
heavy, in spite of its inherent delicacy and softness. This is the case in m. 9, where a lower
bass note and the octave in the right hand allow for a forte climax on f''', the highest note of
Mozart's piano. A crescendo in m 8 towards the fourth beat f''' is not possible on the
fortepiano, as the treble will be overplayed. Besides, the metrical structure points to a
lightening of the (upbeat) fourth beat. All of bar 8 must therefore be slightly diminuendo,
while maintaining direction towards the next downbeat. The m. 9 climax is relative to this
diminuendo and the forte will therefore not be extremely loud.
The modern piano allows for forte in the middle to high and highest registers; as a result,
since the early 19th century a style has developed in which high notes can be louder. The
1893 edition by Sigmund Lebert (1821-1884) of Mozart's K.333 clearly shows this change in
style. On the (as good as as) modern piano with a rich treble extending some 1,5 octave higher
upwards than Mozart's, and within the romantic aesthetic of the time this would sound
entirely natural. Besides the crescendo in m. 8 Lebert often indicates a crescendo to a high
note, even where this so-called highpoint falls on a fourth beat (as in mm. 1 and 6). This
illustrates the romantic aesthetic that most modern pianists go by today.
34
Mozart. Sonata in Bb, K. 333, first movement (1783).
Sigmund Lebert, ed. Schirmer, 1893.
Carl Czerny (1791-1857), Beethoven's most important student, intimate with all Beethoven's
keyboard works and the author of several important treatises among which his monumental,
four volume Vollständige theoretisch-practische Pianoforte-Schule op. 500 (1839) shows
himself to be a product of his age when he wrote that any ascending passage needs a
crescendo. However, in the illustrating example the highest note is written after the downbeat
(in both hands, i.e. in m. 3 as well as in m. 7) and therefore lifted, This passage does not have
any moderating slurs; harmonically the messa di voce is uninterrupted. This means that in this
passage, including the crescendo towards the highpoint, Czerny is not departing from the
earlier practice.36
Czerny Pianoforte-Schule (1839)
36
Carl Czerny. Vollständige theoretisch-practische Pianoforte-Schule, von dem ersten Anfange bis zur höchsten
Ausbildung fortschreitend. In 4 Theilen. op. 500. Part III: “Von dem Vortrage.” Vienna: A. Diabelli, 1839/Facs.
Repr. of 2nd ed. Vienna, 1846: Wiesbaden: Breitkopf und Härtel, 1991. Ch. 1 Von der Anwendung des
crescendo und diminuendo, §2, p. 11. See also the Czerny example in my paragraph on messa di voce (Ch. IV
§4), where Czerny does favour an off-beat highest note as expressive climax and loudest note.
35
The great Fantasy in c minor K. 457 (1785) by Mozart is in many ways an extreme work. This
certainly holds true for the dynamics. While the passionato f major passage from m. 56 starts
p, the active left hand allows the right hand to crescendo towards m. 61. The climactic c''',
however, has an ornament to make the forte possible without forcing the tone. Measure 64 is
marked f again; however this is the first note of a crescendo development from mm. 64-73
and therefore not too loud yet. The crescendo is mostly made possible because of the
descending left hand basses and the Alberti Bass activity.
In the passage following, Mozart creates the largest climax in any of his solo or duo keyboard
works. Keeping in mind the delicacy of the treble, it does not come as a surprise that this
climax takes place in the bass, in fact on the lowest note of Mozart's piano FF (doubled by the
octave) in m. 82 and repeated, preceded by an extraordinary descending octave passage in the
right hand (not to be divided over two hands!) even more powerfully in m. 85.
Mozart. Fantasy in c minor, K. 475 (1785). mm. 56-82.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
Beethoven, using the same five octave piano as Mozart until ca 1800 and therefore bound to
the same dynamical restrictions in the treble as Mozart, nevertheless wrote extremely
spectacular and loud passages in the treble. In order to not overplay the instrument he had to
revert to the same techniques as Mozart.
36
Beethoven. Sonata in C# minor, op 27 no 2 'Mondschein' (1801). mm. 20-40.
First edition, Cappi, n.d.[1802]
The Alberti Bass starting in m. 21 (the second measure of this excerpt), twice as fast as
Mozarts, provides a powerful agitato energy on which the right hand can float. As soon as the
crescendo starts the right hand is written in octaves. When sforzandos are indicated, the bass
descends lower and lower, allowing for more forte. Modern performance has it that this
melody can be performed as a lyrical melody, with a rather suppressed left hand. Indeed, the
left hand easily gets too loud on a modern instrument; but on the five octave Viennese
fortepiano the energy and tension arising from a loud and driving left hand, almost (but not
quite, of course) overwhelming the right hand is the key element of the agitato. Making this
passage essentially lyrical instead of rhythmical and agitato changes the character fatally.
In the next phrase, the forte chords with E''' as the highest note (the second highest of
Beethoven's piano) the large chord accommodates sufficient forte.
Beethoven. Sonata in C# minor, op 27 no 2 'Mondschein' (1801). mm. 46-50.
First edition, Cappi, n.d.[1802]
Besides the five part chord itself, these chords may (and probably must) be played with the
pedal (a knee lever on Beethoven's piano), interestingly enough called 'forte pedal' in the early
days of its existence; much like Beethoven himself prescribes in the opening of this
movement where he asks for the ascending arpeggios to be played with the dempers (i.e.
without using the pedal) and the sf chords without the dempers touching the strings(m. 1-2).
37
Beethoven. Sonata in C# minor, op 27 no 2 'Mondschein' (1801). mm. 1-4.
First edition, Cappi, n.d.[1802]
On string instruments there was a different attitude already earlier on, as the Méthode de
Violon (ca. 1793) by Baillot, Rode and Kreuzer shows: towards the high notes there can be a
crescendo; towards the low notes a dim.
Baillot, Rode and Kreuzer. Méthode de Violon (ca. 1793).
Friedrich Wieck (1785–1873) mentions the habit of singing out on high notes in his treatise
Klavier und Gesang (1853). It is not entirely clear whether he doesn't approve of louder high
notes as a matter of principle, or only when they are forced too much: 37
Do you not think that the natural feeling of an educated and unprejudiced
ear—not often, it is true, the ear of the composer and singing master—is
every moment hurt by the frequent trembling of the voice, the excessive
forcing of high notes (where your chest register is pressing already into the
Falsetto)...
Every classical piano work bears witness to this style feature based on the register. Of course,
good musicians understand that high notes are expressive. But romantic performance practice
asks for that expressivity to be performed louder; classical performance generally asks
(excepting the climactic passages such as mentioned above) for the high notes to be
performed delicately, thus softer.
In her book on Leschetitsky's teaching, Marie Prentner showed that this high romantic
principle was already engrained into the style of the time (1903):
Rule 1. An ascending series of notes demands a crescendo, its highest note the strongest
accent, the descending series of notes a decrescendo.
Rule 2. Of two notes the higher is the stronger.
Rule 3. Of two notes the longer one is the stronger.38
37
Friedrich Wieck. Klavier und Gesang, 1853 / Piano and Song. Transl. Mary P. Nichols, 1875.
Marie Prentner. Leschetitsky's Fundamental Principles of Piano Technique. Part II, Chapter IV. The Melody
and its Rules. Philadelphia: Presser, 1903 / Reprint New York: Dover, 2005.
38
38
5. Touch
In Beethoven's Sonata op 81a, Les Adieux, the 1953 Henle Urtext edition gives no wedges
anywhere, based on the idea that they were too difficult to distinguish from the dots in
Beethoven's rather flexible notation. (This decision has been reversed in the more recent
Henle editions). Moreover, in measures 19 and 20 of the second movement the early Henle
Urtext gives continuous dots for the left hand, where originally there were just wedges on the
first five left hand figures.
Beethoven Sonata op 81a, Les Adieux. (1809-10). Second movement mm. 19-20
Henle Urtext (1953)
Beethoven Sonata op 81a, Les Adieux. (1809-10). Second movement mm. 18-21
First edition, Leipzig: Breitkopf & Härtel, n.d.[1811].
The crescendo in m. 18 leads to a highpoint on the first beat of m. 19, indicated by a sf.39
The diminuendo that follows leaves no doubt as to the dynamics in this passage. However, the
dynamics is achieved not only by actual lessening of the volume, but also by the fact that the
left hand softens its touch.40
An even better example of this type of 'energy' indication of dynamics is the passage mm. 3744 in the third movement, again not faithfully represented by the 1953 Henle Urtext.
Beethoven’s intentions are perfectly clear when the indications of accentuation, dynamics and
pedaling are seen as having one and the same intention, namely to lessen the energy, thus
resulting in a dimuendo.
Beethoven Sonata op 81a, Les Adieux. (1809-10). Third movement mm. 37-44.
Henle Urtext (1953)
39
Possibly the placement of this sf is more subtle than Henle likes to think, and meant for the second sixteenth
note in the right hand which then would supercede the highpoint of the first beat.
40
In the corresponding passage mm. 35-36 in the first edition there are only wedges on the first left hand figure.
Also here Henle has extended them to all eight left hand figures.
39
Beethoven Sonata op 81a, Les Adieux. (1809-10). Third movement mm. 37-44.
First edition, Leipzig: Breitkopf & Härtel, n.d.[1811].
The first four quarter notes are marked ff, sf and have a wedge. The next four (in the same
pedal) are nominally still ff but lack the wedge and the sf. The next four (mm. 40-41) have just
the ff and the wedges but no sf. In the new pedal these notes will therefore be less loud. The
last four have no wedges and no sf. The net result is an 8 bar diminuendo, leading to the
following p passage (although the character change makes a clear dynamical break logical).
The Breitkopf & Härtel Ludwig van Beethovens Werke of 1862-90 already gave the markings
exactly as Henle did in 1953. Interestingly, Schenker adapted in 1918 the markings in this
passage in such a way that it could also result in a slight diminuendo, by adding dots
(indicating relative lightness, compared to the wedges) to mm. 39-40 and leaving mm. 43-44
without dots or wedges; but he added sf to all notes, intending a forceful performance
throughout.
Beethoven Sonata op 81a, Les Adieux. (1809-10). Third movement mm. 37-44.
Schenker: Universal Edition, 1918-21. Repr/Dover, 1975
The often scorned Martienssen in the 1920 Peters edition, however, gives by far the best
representation of this passage, leaving open (through the use of brackets) the possibility of the
loss of energy resulting in a diminuendo.
Beethoven Sonata op 81a, Les Adieux. (1809-10). Third movement mm. 37-44.
Martienssen: Peters, ca.1920.
40
IV. Interpreting dynamics
When in early repertory dynamics are lacking or notated very generally, a performer will have
to come to an understanding of what this means. For a classical score the point of departure
should be that the notation is general, incomplete and in need of interpretation. As discussed
above, it is possible to complete a general dynamic scheme with refined details, based on an
understanding of the musical language. Dynamics may be implied in the notation or in the
way the music sounds, moves and develops. Other elements of the musical notation may have
an effect on dynamics and even the instrument itself has a say in it.
In this chapter we will look at dynamics that are notated but may have a meaning beyond
what is expected. Dynamics may alter other properties of performance, such as tempo rubato,
accentuation and even meter. Some seemingly clear markings such as a forte, a crescendo, or
a diminuendo may have an additional meaning.
1. Tempo rubato
Türk identifies a specific type of Tempo Rubato in his treatise Klavierschule (1789), which is
indicated with dynamics:41
Außer der angezeigten Bedeutung des Tempo rubato versteht man unter
diesem Ausdrucke zuweilen auch nur eine besondere Art des Vortrages,
wenn nämlich der Accent, welcher den guten Noten zukommt, auf die
schlechten verlegt wird, oder mit andern Worten: wenn man die Töne auf
dem schlechten Takttheile usw. stärker vorträgt, als diejenigen, welche in
die gute Zeit des Taktes (oder einer Note) fallen, wie in diesen Beyspielen.
Türk Klavierschule (1789), p. 375.
Exactly this type of passage appears in Mozart's Sonata in c minor, K. 475, second movement.
It is just one of several passages which point to Mozart's extensive use of rubato. There is
little doubt that this is an integral part of Mozart's style (as well as, in my opinion, of the
entire classical style) - but not always does he indicate this in a way which makes it clear to
modern interpreters.
41
Türk Klavierschule. Ch. 6 Part V, §72. p. 375.
41
Mozart. Sonata in c minor, K. 457, second movement. (1784).
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
The 'syncopated' notation in m. 19 (also treated by Türk in his chapter on tempo rubato)
prepares the listener for the rubato which is the natural result of the f-p-f-p markings in m. 21.
The relative lengths of the forte and piano notes must be different and therefore indicating the
dynamics is not primarily a way to govern the loudness and softness, but the flow of the
passage.42
Solche Stellen würden schlechte Wirkung thun, wenn man sie genau nach
der bestimmten Geltung der Noten (taktmäßig) spielte. Die wichtigern
Noten müssen daher langsam und stärker, die weniger wichtigen aber
geschwind und schwächer gespielt werden, ungefähr so, wie ein
gefühlvoller Sänger diese Noten singen, oder ein guter Redner die Worte
dazu deklamiren würde.
Rubato notation by way of dynamics is not limited to the 'romantic' Mozart (as he was
sometimes called by his contemporaries) but also appears in Haydn, whose style is often
perceived today as more austere and 'classical' - a term which has been defined by generations
of musicians since the mid 20th century as 'strict in tempo'.
42
Türk Klavierschule. Ch. 6 Part V, §72. p. 371.
42
Haydn. Sonata in c minor, Hob XVI/20. (1771) First movement mm. 8-15.
Martienssen: Peters, ca.1920.
The succession of dynamics in m. 14, exactly as a few years later practiced by Mozart and
again five years later described by Türk, can not be performed other than with a subtle tempo
rubato which provides an expressive variant of the gesture in m 11.
2. Forte markings
Often, Beethoven and other classical and early romantic composers use a string of forte or
sforzando markings to indicate a succession of loud notes, together building to a climax.
These should be interpreted as recurring climactic notes and the f therefore functions as a type
of accent, perhaps without the sharp attack that a sf would seem to indicate.
In the Sonata in C minor, op 10/1, between mm. 78 and 90 the phrase develops to a climax
twice by way of three counterbeat sforzandos, resulting in a climactic ff on the downbeat (in
this edition the first starts line 1, m 8). The fortissimo marking is not absolute, but only
indicates the climax. In particular the second one (m. 90, here line 2 m. 9), solo on a high note
without bass support, needs to be dealt with carefully on any instrument, but particularly on a
five octave fortepiano.
Beethoven. Sonata in C minor, op. 10/1. 1796/98. First movement, mm. 71-103.
First edition: Eder, n.d.
The relatively delicate highpoint of m. 90 leads to a better, more powerful climax in the bass
in m. 93, immediately followed in the same bar by a right hand sf on the third beat. This last
accent is in fact a climactic note inside a climax and could therefore be called sublime, the
highest order of climax obtainable.
43
Beethoven Sonata op 13, Pathétique. First movement mm. 108-135.
First edition, Vienna: Eder, n.d.[1799].
In the first movement of the Sonata Pathétique the f markings in mm. 127-129 (3rd line, mm.
8-10) and the resulting climactic ff should be interpreted as recurring climaxes, functioning as
a crescendo to the ff in 130.
The spectacular climactic high note in m. 58 of the third movement of the Pathétique Sonata
is prepared by two sforzandos in the left hand (mm. 54-55) and a crescendo (starting in the
previous predominant dynamic p, this time without sf) as well as an ascending figuration to
the highest note of Beethoven's piano, f'''. This high note is all but delicate - its loudness is
guaranteed by the low GG-G octave, effectuating a spread of almost the complete five octave
compass of the instrument. The hands spreading provides an extra indication for the climactic
effect of m. 58 marked ff, which in this case must be understood as an indication of the
highpoint. This means that the descending scale (the hands converging again) in mm. 59-60
must be diminuendo, while the character demands that this be done without loosing energy.
As a result the sf on the dominant of m. 60 is in fact an accent in the context of a much lower
dynamic than ff, rather than a final note even louder than the high note climax as modern
performance practice advocates with the argument that 'no diminuendo is notated'.
Beethoven Sonata op 13, Pathétique. Third movement mm. 53-61.
Henle Urtext (1952).
The arrangement made by Hummel (1778-1837) of Beethoven's Fifth Symphony, op 67 for
piano, flute, violin and cello (premiered in 1808) shows how a forte indication could be used,
besides as a high point, also to identify a new entry, or a voice which needed to be brought out.
In the last movement, a lyrical passage in the development starting on the upbeat to m. 480 (in
44
this edition line 1, m. 3) is marked f by Hummel. The manuscript copy as well as the first
edition also give this forte. Hummel furthermore gives fortes to three out of four entries of the
winds with the four note theme Bb-A-Bb-C, first heard in the upbeat to m. 480 in bass:
and in all ensuing theme entries. This means that the f in m. 482 (end of line 1) applies to the
bass again (C-Bb-C-D), the ff to the middle notes Db-C-Db-Eb of the right hand chords
starting on the upbeat to m. 486 (line 2, m. 4; ff because of the weaker fortepiano treble), and
one more forte on the upbeat to m. 492 (line 3, m. 5, F-E-F-G) before the più f in 495 (line 4,
m. 2).
Beethoven Symphony no 5 in C minor, op 67 (1804-1808). Fourth movement mm. 478-504
Arr. J.N.Hummel for Piano Quartet: Schott [n.d.]
However, in the manuscript copy as well as the first edition of the symphony there is but one
forte indication on the upbeat to m. 478 (first ms. page, upbeat to bar 4) and one f indication
for the trombones (ff for the bassoon) on the upbeat to m. 486 (second ms. page, upbeat to bar
2). There is nothing else until the più f in m. 495. Clearly, in his arrangement Hummel wanted
to draw attention to the new theme entries. Whether he felt that this was necessary because of
the fact that the piano cannot make them stand out well enough for lack of a contrasting color
(the brass more easily standing out from the orchestra sound) or whether it was a habit of
compositional clarity to write f over prominent voice entries remains open to debate.
45
Beethoven Symphony no 5 in C minor, op 67 (1804-1808).
Fourth movement mm. 477-480. Copyist's manuscript [n.d.]
Beethoven Symphony no 5 in C minor, op 67 (1804-1808).
Fourth movement mm. 485-488. Copyist's manuscript [n.d.]
3. Counter accentuation
In his discussion of the correct performance of the slur, C.P.E. Bach mentions that the slur can
cause an accent against the metrical beat: "in other cases the note on which the slur starts
receives this pressure:"43
§18. Die Noten, welche geschleift werden sollen, müssen ausgehalten
werden, man deutet sie mit darüber gesetzten Bogen an. Dieses Ziehen
43
Bach Versuch. Part I (1753), Ch. III: Vom Vortrage. p. 94, §18. See also my Ch III, §3, The correct
performance of the slur".
46
dauert so lange als der Bogen ist. Bey Figuren von 2 und 4 solcher Noten,
kriegt die erste und dritte einen etwas stärkern Druck, als die zweyte und
vierte, doch so, daß man es kaum merket. Bey Figuren von drey Noten
kriegt die erste diesen Druck. Bey andern Fällen kriegt die Note diesen
Druck, wo der Bogen anfängt.
In his chapter on bowing, Leopold Mozart shows a four bar passage for which he proposes no
less than 34 different bowings. In particular nos 11 and 34 are interesting, as these bowings
start off beat; in order to make his intentions clear, Mozart indicates the up and down strokes
underneath these examples.44
Leopold Mozart. Versuch (1755).
In the next paragraph, Mozart explains how these extraordinary bowings change 'the entire
performance' when the heavier first note of each slur may fall on a weak beat.45
Wenn nun in einem musikalischen Stücke 2. 3. 4. und noch mehr Noten
durch den Halbcirkel zusammen verbunden werden, daß man daraus
erkennet, der Componist wolle solche Noten nicht abgesöndert, sondern in
einem Schleifer singbar vorgetragen wissen: so muß man die erste solcher
vereinbarten Noten etwas stärker angreifen, die übrigen aber ganz gelind
und immer etwas stiller daran schleifen. Man versucht es in den vorigen
Beyspielen. Man wird sehen, daß die Stärke bald auf das erste, bald auf das
andere oder dritte Viertheil, ja oft sogar auf die zwote helfte des ersten,
zwoten oder dritten Viertheils fällt. Dieß verändert nun unstreitig den
ganzen Vortrag [....]
Türk demonstrates how the rule of good and bad beats in a bar can be superseded by the
composer. The composer only has to indicate the dynamics to make sure that the performer
understands correctly.46
As we have seen above, Türk agrees with C.P.E. Bach and Leopold Mozart on the fact that
slurs start louder and sing out gradually softer, as well as having a slightly lengthened opening
note and shorter/faster following notes (as do all treatises which deal with the subject).
Combined with his readiness to alter the good and the bad beats this leads to a surprising
conclusion, namely that the regular accentuation in a bar may be changed and that counter
accents can be introduced by the start of a slur - much like the examples 11 and 34 of Mozart.
Türk particularly mentions this effect, illustrated in his example g and h: "in example g) this
44
Mozart Versuch. Ch. VII, Part 1. pp 131-136, §§ 19-20.
Mozart Versuch. Ch. 7, §20, p. 136.
46
Türk Klavierschule. Ch 6, Part II, §13.
45
47
slight stress [...] falls therefore on the with + designated bad notes, in h) on the f sharp, d, b,
etc." 47
Bey den Tonen, welche geschleift werden sollen, läßt man den Finger so
lange auf den Tasten liegen, bus die Dauer der vorgeschriebenen Noten
völlig vorüber ist, daß also auch nicht die kleinste Trennung (Pause)
entstehe. Durch den längern oder k¨rzern Bogen bestimmt der Komponist,
wie viele Töne an einander geschlieft werden sollen. S. B.
Bey a) werden folglich alle acht Töne, bey b) vier und vier usw. geschleift. Man
merke hierbey noch, daß die Note, über welcher der Bogen anfängt, sehr gelinde
(kaum merklich) accentuirt wird. In dem Beyspiele g) fällt also dieser gelinde
Nachdruck, (wider die sonst zu befolgende Regel) auf die mit + bezeichneten
schlechten Noten, bey h) auf fis, d, h, u.s.w. Durch die Bezeichnung bey k) will
man andeuten, daß zwar alle Noten geschleift werden müssen, doch soll
der erste, dritte, fünfte und siebente Ton sehr schwach markirt werden.
This feature of the slur has not been fully understood in present day music life. While the slur
nowadays stands exclusively for legato and in a larger perspective for lyricism (expressivity),
or its meaning sometimes is played down to nothing more than a playing technique, the
additional features of the slur signifying (harmonic or melodic) stress, the ensuing relaxation,
tempo rubato or a pronounced rhythmic profile may cause us to understand pieces entirely
differently.
The Scherzo of Beethoven's Sonata in G major, op 14/2 has an eccentric and striking
articulation.
Beethoven. Sonata in G major, op 14/2, Scherzo (1798-99)
Henle Urtext (1952)
47
Türk Klavierschule. Ch 6, Part III, §38. See my chapter II,3,b. The fact that the beginning of a slur has a slight
accent also may shape a rhythmical profile, even under a larger slur as in k).
48
If the slurs are performed according to 18th c performance practice, the meter of the first two
bars is not 3/8 but 2/8; the barline is perceived to fall after the first eight note of bar 1, and so
on. The last note of bar two, a surprising dissonance not belonging to the key of G major, is
seperated from the notes before and after, with a dot to indicate its lightness. This note
interrupts the 2/8 pulse. Next, when understood in the 2/8 meter, the first beat of bar 3 is
perceived to be lengthened and slurred to include the downbeat of bar 4. As there can be no
accent under the slur other than the first note, the 2/8 meter can not be continued. The listener
suddenly understands that he has been fooled by the 2/8 meter and that it really should be a
3/8. The fact that bar 4 has no real downbeat makes it possible to play this joke again in bars
5-8: the listener might still have heard four gestures, even if the downbeat counting starts on
the first (in reality upbeat) note of the piece. In short, this Scherzo is a rhythmical joke; it is a
scherzo in the real sense of the word, befitting a composer like Beethoven. But only if the
slurs are performed in the way described by every 18th c treatise.48
4. Messa di voce
In his treatise Die Wahre Art das Pianoforte zu spielen (1799), Johann Peter Milchmeyer
(1750-1813) - like all his contemporaries - describes the correct performance of the slur in the
same way as Leopold Mozart did so emphatically almost half a centry earlier, in 1755. There
is an interesting addition to his description, however: in slurs longer than 4 notes the
performer may add a crescendo-decrescendo.49
Bei zwei gebundenen Noten ist die erste, die stärkste, die zweite die
schwächste, und bei drei gebundenen Noten, (wenn es nicht Triolen sind)
wird die erste am stärksten, die zweite schon etwas schwächer, und die
dritte am schwächsten gespielt, bei vieren findet eben diese Abstufung von
stärkern und schwächern, von der ersten bis zur vierten Note statt. In
gebundenen Gängen von 6 oder 8 Noten, die nicht aus Läufen, aus auf- und
abgehenden Noten bestehen, onder hin- und hergehende Gänge sind, macht
man die mittelsten Noten etwas stärker, und die ersten und letzten
schwächer. Diese Ausdruck wird so bezeichnet, < >
Domenico Corri (1746 - 1825) treats the subject of 'Messa di voce', which he calls 'the soul of
music' in The Singer's Preceptor, or Corri's Treatise on Vocal Music (1810) with special
attention.50
48
I would like to point out that in addition to this, the high notes in mm. 10 and 12 fall on the second eight note
and are therefore to beplayed p rather than f.
49
Milchmeyer Wahre Art. p. 46.
50
Domenico Corri (1746 - 1825). The Singer's Preceptor, or Corri's Treatise on Vocal Music. London: Chappell,
1810. Lesson I, p. 14. I am indebted to Arturo Peres Fur's paper for the Fortepiano Literature Class, 2012-2013
season, Royal Conservatory The Hague.
49
Corri. The Singer's Preceptor (1810)
In the Third Prerequisite for good singing (p. 65) Corri explains how the messa di voce is
applied to a whole phrase. From his examples it appears that also to Corri it is important that
the slur never ends with a crescendo.
Corri. The Singer's Preceptor (1810)
Johann Adam Hiller's Anweisung zum musikalisch-zierlichen Gesänge (1750) was already
quoted above. Like Corri, Hiller mentions the messa di voce as one of the beauties in
singing.51
Diese alle muß ein guter Sänger in seiner Gewalt haben, und mit ihnen
bringt er nicht allein das gute Portamento di voce, sondern auch noch eine
andere Schönheit, ein allmählicher Verstärken und Schwachen des Tons,
das sogenannte messa di voce hervor. Dieses kann vom pianissimo bis zum
fortissimo getrieben, und so auch wieder bis auf jenes zurück gebracht
werden.
Obviously, the piano can not produce a messa di voce on a single tone like Corri describes.
However, many composers have written just that. In long slurs, also on the keyboard a messa
di voce is in order. Mozart uses relatively long slurs in the Sonate in F major, K. 332, both in
the first (m. 68-69) and second movement (m 3). In each case a crescendo is possible during
the first half of the slur, but the end of the slur must still have a diminuendo.
51
Hiller Anweisung. Ch II, §8.
50
Mozart. Sonata in F major, K. 332, first movement. 1783.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
Mozart. Sonata in F major, K. 332, second movement. 1783.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
Beethoven generally wrote more and longer slurs than Mozart. In fast pieces with short meters,
Beethoven's slurs easily stretch four bars. Without exception, the relative stress of the phrase
under the slur produces a messa di voce. In this example from Sonata op 2/2, the messa di
voce in m. 27 is indicated fp (a louder note in a piano context) by Beethoven. Two bars later
the left hand imitation is written down differently. As the left hand high point is even more
important, the climax is indicated by sfp. Here, the slur is broken: the heavier new beginning
of the slur produces the emphasis which creates the highpoint.
Beethoven. Sonata in A major, op. 2/2. 1795. First movement, mm. 18-33.
Henle Urtext (1952).
What is crucial in understanding this instrumental messa di voce is that, in case there is a need
for an expressive highpoint in the middle of the slur, the slur still always has a diminuendo at
51
the end. If these basic requirements are applied, a new understanding of familiar passages
may arise.
The slur from measure 1 to 2 in the opening of Beethoven's Sonata in c minor, op 10/1 has a
slight stress in the middle, namely on the downbeat of measure 2. The upward gesture ending
on Eb, almost the highest note of Beethoven's piano, can not have a crescendo to the top as
the highest note would be overplayed; the slur must end with a diminuendo in order to end as
lightly as indicated by Beethoven himself (dots on both E flats). The net result is a messa di
voce with the downbeat in the middle as the focus point.
Modern performance sometimes turns the upward gesture into a crescendo rocket gesture,
with the loudest note on the downbeat of m. 3. Besides ignoring the lightness indications on
the high notes and the inherent shape of the slur, this creates yet another problem. The larger
pulse of the piece is the result of the placement of good and bad bars. In this case, in spite of
the forte opening chord, the good bar is bar 2 and therefore also on bars 4-6-8-10-12. The four
bar (inconclusive) ending of the phrase with the focus point in the third bar of those four leads
to a heavier bar 15. By emphasizing m. 3, diction problems arise everywhere, like in in mm.
9-10 where the longer slur is performed with a messa di voce as the dominant chord on the
downbeat of m. 10 needs to be emphasized.
Beethoven. Sonata in C minor, op. 10/1. 1796/98. First movement, mm. 1-50.
First edition: Eder, n.d.
The second theme starting on m. 32 (in this first edition: line 3, m. 9) has two long slurs. The
four bar left hand slur starting in m. 33 (line 3, m. 10; for this discussion leaving aside the
extra bar in bar 32) focuses on the third bar, where the dominant appears. To ensure the
correct focus, Beethoven has tied the Eb in the middle voice. The right hand, focussing in
principle at the same place as the left hand, plays a syncopated downbeat before the third bar.
This is the note that receives most of the dynamical stress in the right hand. This messa di
voce is carefully notated by Beethoven; after the syncopated E flat the right hand sings out
with a diminuendo. It follows that the higher f in m. 34 (on the weak beat 2) is not to be
played as the highpoint of this gesture.
The next four bar gesture has the same structure. The third gesture of this phrase, starting in m.
41 (line 4, m. 5) is twice as long. In order to make sure that the correct messa di voce is
applied, Beethoven indicated the highpoint with a hairpin towards m. 6 of the gesture. Again,
52
the highest notes (the A flats in m. 41-42 and the B flat in 42) are not to be played as
dynamical highpoint.
Czerny, the greatest expert on Beethoven, also devotes a paragraph on messa di voce. He
shows two different ways to indicate it. Each four bar slur receives a messa di voce. In the
first phrase Czerny there is a diminuendo from the first beat of bar 4; the high g in bar 3 is
therefore not the dynamical highpoint, but the entire third bar it is the expressive highpoint in
which the high g is the relatively loudest note. It is interesting to note that under the second
slur the highest note on the third beat is indeed marked as the dynamic climax of the gesture.52
Czerny's final remark may serve as the famous last words for this paragraph: "It goes without
saying that the performer must apply this expressivity himself, when the composer has not
indicated it."
5. Crescendo as rinforzando
In modern performance practice crescendo is interpreted as 'getting louder until the next
marking intervenes'. This has, all by itself, created a style of playing that is full of dynamical
surprises, even where the character of the piece does not seem to ask for it.
The 18th century treatises have, without exception, definitions that do not coincide with this
view. From their pages arises a much different picture, namely one in which a crescendo may
ask for an increase of volume just for the notes under which it is written. This means that a
crescendo marking may function - and indeed seems to do so often - as a local rinforzando.
Leopold Mozart (1755) is clear in his instruction:53
Crescendo, wachsend, will uns sagen daß wir bey der Folge der noten, bey
welchen dieses Wort stehet, mit der Stärke des Tones immer anwachsen
sollen.
52
Carl Czerny. Vollständige theoretisch-practische Pianoforte-Schule, von dem ersten Anfange bis zur höchsten
Ausbildung fortschreitend. In 4 Theilen. op. 500. Part III: “Von dem Vortrage.” Vienna: A. Diabelli, 1839/Facs.
Repr. of 2nd ed. Vienna, 1846: Wiesbaden: Breitkopf und Härtel, 1991. Ch. 1 Von der Anwendung des
crescendo und diminuendo, §4, p. 13.
53
Mozart Versuch. Ch. I, Part III. pp 131-136, §27. p. 52.
53
This definition and the following (correct) translation by Editha Knocker54 has never been
fully appreciated. "Crescendo: means increasing, and tells us that the successive notes, where
this word is written, are to increase in tone throughout" (underlining mine). Rather than
interpreting this in the usual way I believe that this means that the increase of volume is not a
continuous process, but takes place only where the word is written.
Türk (1789) mentions the fact that a performer sometimes may not know exactly how long the
crescendo is to last: "Since at the word crescendo one can not exactly know how long the
gradual increase must last, some would write a f (forte) under the note which is to be played
with full loudness."55
cresc. - crescendo, wachsend, zunehmend; (stärker werdend;) wird auch
durch dieses Zeichen: < angedeutet.
Da man bey dem Worte crescendo nicht genau wissen kann, wie weit
dieses allmähliche Anwachsen gehen soll, so pflegen Einige unter die
wieder mit völliger Stärke zu spielenden Noten ein f (forte) zu schreiben.
A poco a poco crescendo fina al (il) forte, nach und nach wachsend zur
Stärke.
With this it seems that Türk acknowledges that the crescendo may have an open ending; he
says that some composers mark a forte to indicate where the crescendo leads. Reversely,
without a marked forte after a crescendo a continuous crescendo may not be asked for. And
indeed the next entry takes away all doubt: only when at the end a forte is notated, the
crescendo continues until that moment. In order to ensure correct understanding the composer
will have to make his intentions clear, for instance by this rather unusual and lengthy marking
"a poco a poco crescendo fino al (il) forte". It is true that this marking is rare in classical
scores; but at least it demonstrates that in order to make the crescendo last, the composer
needs to indicate some goal. When this is lacking the crescendo can and often must be
interpreted as a local effect, like Leopold Mozart instructed.
In 1797, Milchmeyer's definition of crescendo seems to coincide with modern interpretation:56
Crescendo (cr.) bedeutet, daß man jede Note vom Anfange des Ganges bis
zum Ende immer stärker, als die vorhergehenden spielen soll.
But the example that accompanies this passage leaves no doubt as to the nature of this
crescendo: only when it leads to an actual forte the crescendo is maintained. By the way, there
is no slur in this example which could moderate the continuity of the increase.
Milchmeyer. Die Wahre Art das Pianoforte zu spielen (1797)
54
Leopold Mozart. A Treatise on the Fundamental Principles of Violin /Translated by Editha Knocker
1948/reprint of 2nd ed. Early Music Series no. 6. Oxford: Oxford University Press, 1986.
55
Türk Klavierschule. Ch 1 Abschnitt 5, §79, p. 117.
56
Milchmeyer Wahre Art. Ch. IV: Vom Musikalischen Ausdruck, p. 50.
54
We must conclude from these definitions that, at least before 1800 (but I believe in many
cases after 1800 as well) a crescendo marking is first of all a local event, rather than an
unrelenting process of increasing the dynamic level. Only when leading to a loud note (f, ff, sf,
or another type of accent) the crescendo may open up to a new register. In all other cases
crescendo functions primarily as a rinforzando.
A most striking and often misunderstood example of this 'rinforzando' use of the crescendo is
found in one of the great masterpieces of the classical period, Mozart's Sonata in C minor,
K.457, second movement. Besides being a showcase of thematic variation, ornamentation and
rubato notation, this magnificently detailed and meticulously notated movement is replete
with dynamic indications, the most striking of which are the crescendo markings in mm. 14,
24, 27, 31, 33, 54, where in each case the crescendo is followed by a p.
Mozart. Sonata in C minor, K. 457 (1784). Second movement, mm. 13-16.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
First of all, the first three eight note slurs starting on the third beat of the first bar do not have
the crescendo. As they have a clear p marking on the first note, annulling the normal heavier
start of the slur, the most logical expressive interpretation is a messa di voce, the middle being
so expressive. This is therefore also the basic shape of the figure with crescendo starting on
the second beat of m. 14. The crescendo being local, it can now be interpreted as reinforcing
the expressivity of the middle of the slur, leaving room to diminish the last notes of the slur. It
is an open question whether the diminuendo on the last notes of the slur actually go down to
the p level, of whether there is still a slight break.
When we look at the concurrent passage in m. 54-55, the agitato nature of the phrase is
enhanced dramatically by the addition of three extra crescendos on the first three figures.
Mozart. Sonata in C minor, K. 457 (1784). Second movement, mm. 54-55.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
It remains a matter of personal taste and temperament whether the crescendo implies a certain
amount of rubato, adding to the agitated nature of the passage. Besides rubato within each
55
gesture, the added agitato and possible drive of the five crescendo gestures in mm. 54-55
makes mancando in the second bar logical. These decisions are the domain of the performer.
It is not an option, however, to keep increasing the volume on each gesture until the next beat
and create a succession of subito p moments.
Mozart. Sonata in C minor, K. 457, second movement. 1784. mm. 23-26.
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
Like at the end of m. 14, in m. 25 the crescendo leads to a forte. Similar crescendos in this
movement happen in mm. 3, 7, 14, 19, 23, 34, 35, 36, 37, 43, 47, 55. In each and every case,
the crescendo is notated under a slur.
Turning to m. 25, the cresc. marking may to some seem to point to a continuous crescendo in
the seven notes under the slur (m. 25, fourth beat). However, if the crescendo is understood to
be local, a rinforzando is in order on the first half and a (minimal) diminuendo at the end of
the slur, resulting in a messa di voce. The forte, moreover, starts in each hand on the second
note of bar 26, meaning that the downbeat of m. 26 does not have to be loud.
The interpretation suggested by these elements – the crescendo acting as a rinforzando, the
slur needing a diminuendo but possibily having a messa di voce – is that the interpretation of
the slur is exacly like the one described in m. 14. Nevertheless, the final dynamic of the slur
as well as the energy (aided by the left hand repeated notes) may be higher than in m. 14.
Given the rising tension from m. 24, the downbeat of m. 26 is likely to be louder than p. How
much louder is the performer's decision.
Two further determining factors in deciding about the function of the crescendo marking are
whether the word is abbreviated or not and whether there are dots following the word.
While Mozart never uses dots after a crescendo marking, Beethoven does so frequently. These
dots make the crescendo continue without tapering off, making a messa di voce impossible. In
this case there can be no doubt about the target dynamic: either the ensuing forte is reached or
a following p is a subito piano.
In An die Ferne Geliebte most of the crescendo markings are followed by dots. In the third
song Beethoven leaves no doubt about the fact that wants a subito piano on 'meine Thränen'.
56
Beethoven, An die Ferne Geliebte op 98. III: Leichte Segler in den Höhen.
First edition. Vienna: S.A. Steiner & Co., n.d.
The same happens in the fifth and sixth songs:
Beethoven, An die Ferne Geliebte op 98. V: Es kehret der Maien, es blühet die Au
First edition. Vienna: S.A. Steiner & Co., n.d.
A bit earlier in the same song, Beethoven uses a hairpin to make sure that the intended messa
di voce is clear to the performer.
Beethoven, An die Ferne Geliebte op 98. III: Leichte Segler in den Höhen.
First edition. Vienna: S.A. Steiner & Co., n.d.
The transition from the fourth to the fifth song has a crescendo with dots, in order to leave no
misunderstanding:
57
Beethoven, An die Ferne Geliebte op 98. IV and V:
Diese Wolken in den Höhen / Es kehret der Maien, es blühet die Au
First edition. Vienna: S.A. Steiner & Co., n.d.
The few remaining crescendo markings in this cycle without dots following them and without
a forte target dynamic can and probably should be understood as local rinforzando-type
crescendos.
Beethoven, An die Ferne Geliebte op 98. I: Auf dem Hügel sitz ich spähend
First edition. Vienna: S.A. Steiner & Co., n.d.
The crescendo on 'glühend' (second system, first bar) functions as a messa di voce on the
syncopated b natural, resolving with a slight diminendo to the (admittedly also dissonant) c
before the barline. The placement of the 'cresc' in the manuscript is suggestive of this:
Beethoven, An die Ferne Geliebte op 98. I: Auf dem Hügel sitz ich spähend
manuscript (no date)
58
Also in Beethoven's Sonata op. 109 most crescendos are followed by dots. The immensely
refined dynamics can not be notated in a general way anymore, as Beethoven has started
using dynamics in an extremely colourful way.
Beethoven, Sonata op 109. First movement, mm. 1-14.
First edition. Berlin: Schlesinger, 1821.
The cresc in m 4 is followed by dots (which should continue to the target dynamic forte on m.
9; all subsequent editions have correctly rectified this ommision). The forte on m. 9 is
followed by a diminuendo towards p: the left hand follows the decay of the right hand dim. 7
chord, leading to the lowpoint marked p on VI on the forth beat, not subito p . A cresc.
towards the second bar of the Adagio leads to forte. This f marking seems to be printed
underneath the first note of the bar; but the fact that this is a solo melody note without support
of a bass in combination with the powerful character of the seven part chord on the second
eight note, as well as the practical fact that there is no room for the f to be written between the
right and the left hand, makes it plausible that the forte is intended as a subito on the second
eight note. However, there is no p marking on the downbeat of m. 10. The upbeat to bar 11
(third system, first bar) is again marked cresc. and the downbeat of bar 11 is marked piano.
The first edition has this marking on the first two notes of the fourth beat without dots,
leaving room for doubt about whether or not there may be a diminuendo implied (and
therefore a messa di voce) before the barline. In the next bar, with its ornamented upbeat,
there is more space and the dots do appear, leading to the forte in m. 12. From this description
it follows that another way of looking at this passage is possible: the upbeats to mm. 10 and
11 are in fact messa di voces - a view reinforced by the portato, i.e. relatively light marking on
the last two notes of each bar - to a p downbeat and only on m. 12 the dynamics open up to
forte.
A real subito piano is notated in m. 12. The m. 11 cresc. leading up to the barline has dots and
therefore there is no doubt about Beethovens' intentions. The same gesture appears at the end
of bar 12.57
57
The '0' at the end of the bar is a pedal release sign.
59
The Breitkof und Härtel edition (1862-1890) has set a standard for subsequent editions by
'normalizing' the dynamics. Most strikingly the forte marking on bar 10 is assigned to the
downbeat and the crescendo in m. 10 appears clearly on the last two notes of the bar, making
a subito p necessary.
Beethoven, Sonata op 109. First movement, mm. 9-12.
Breitopf und Härtel (1862-1890).
Interestingly, B&H give the pedal release sign, one eight note before the first edition does.
But Beethoven's subtlety found its way into Schenker's Universal edition (1913).
Beethoven, Sonata op 109. First movement, m.12.
Universal, Schenker (1913)
In the tempo primo following the Adagio, the crescendo starting in m. 20 (first line, m. 6) has
no dots and no target forte and therefore implies a messa di voce. The following (m. 26; here
second line, m. 5) has no dots either but leads to a mild sfp.
Beethoven, Sonata op 109. First movement, mm. 15-34.
First edition. Berlin: Schlesinger, 1821.
B&H (1862-90) do add dots to the second crescendo, suggesting that this crescendo does lead
to a louder place. Bülow (Cotta, 1875) does the same and in addition reinforces this
interpretation by adding a diminuendo in m. 23 at the end of the first crescendo.
60
Beethoven, Sonata op 109. First movement, mm. 20-32.
Cotta, Hans von Bülow (1875)
Sometimes a crescendo may be no more than a suggestion, as in the famous last line of the
forst movement of the Les Adieux Sonata, where Beethoven asks for a cresc. on a single long
note. It is not an indication for a messa di voce, as the dots make the crescendo last until the
final chords. The crescendo cannot be realized through a left hand figuration or even a pedal:
it is an imaginary crescendo, perhaps implying a certain stress (making the resolution come a
bit earlier?). Interestingly, this crescendo appears in the manuscript but not in the first edition.
The dots are not printed in the editions of Breitkopf und Härtel (1862-90), Breitkopf und
Härtel (Lamond, 1918), Peters (Martienssen, ca 1920), Universal (Schenker, 1918-21) and not
even in Casella's instructive edition for Ricordi (1920).58
Beethoven Sonata in Eb op 81a, Les Adieux. First movement mm. 249-255.
Henle Urtext (1953).
Beethoven Sonata in Eb op 81a, Les Adieux (1809-10). First movement mm. 248-255.
First edition: Breitkopf & Härtel, n.d.[1811].
6. Decrescendo vs. diminuendo in Schubert
Schubert used both the (abbreviation of the) markings decrescendo and diminuendo often, as
well as closely together within the same piece. Upon careful consideration it seems that for
Schubert these markings had a different meaning. Diminuendo (usually abbreviated to dim.)
always appears at a structural place and seems to always finish a musical thought. Based on
the different ways he uses both markings, this implies loosing energy and if the performers
58
Since the C octave resolves to Bb, it must ultimately have a diminuendo. By absence of the dots in the first
edition, the crescendo could therefore indicate a messa di voce. But this would make the entire marking a mind
game.
61
wishes so, dim. can lead to a slight slowing down. On the other hand, slowing down is
absolutely not an option with decrescendo (decresc.) which appears much more often and can
be found at indiscrimate places in the score. This can also be at the end of a phrase; but
because of the way Schubert often combines both markings there can be no doubt that decresc.
does not allow for slowing down and just serves to bringing down the volume.
Because of this difference in meaning, when used by Schubert in combination dim. always
follows decresc., and never (with one exception, described below) preceeds it.
All of Schubert's works show this feature. One of the most striking (but by no means isolated)
examples is found in the Grazer Fantasy, D. 605a, where the decresc. and dim. signs follow
each other in consecutive bars.59
Schubert. Grazer Fantasy, D. 605a/mm. 157-158
Henle Verlag.
In the last three Sonatas D. 958, 959 and 960 decresc. appears much more often than dim.
Schubert creates large blocks within these enormous structures by the way he identifies and
carefuly finishes musical thoughts with his dim. markings; it is only logical there are not that
many of them. Schubert was careful with dim., it follows that it must be handled with great
care by the performer as well.
Two decresc. markings preceed the dim. at the end of he exposition in the Allegro first
movement of the Sonata in C minor, D. 958. Similar use can be found in the A major Sonata,
D. 959.
Schubert Sonata in C minor, D. 958, i. Allegro, mm. 82-99
Franz Schubert's Werke, Serie X, No.14
Leipzig: Breitkopf & Härtel, 1888 (Epstein).
59
The use of dim. and decresc. particularly in the Grazer Fantasy is the subject of a case study below.
62
It is not surprising that the last movement of the C minor Sonata has no dim. marking
('loosing energy') at all, as one of the defining characteristics of this movement is its great
motorical energy.
Just how careful Schubert was with dim. becomes clear in Sonata in Bb maj, D. 960: in the
entire Sonata there is only one dim. marking: in the second movement, just before the return
of the opening theme. For the rest of the piece he uses only decresc. This has important
implications for the character of this sonata.
Schubert. Sonata in Bb maj, D. 960, ii. Andante sostenuto, mm. 80-94
Franz Schubert's Werke, Serie X, No.15
Leipzig: Breitkopf & Härtel, 1888 (Epstein).
Similarly, in the first of the 2 Marches caractéristiques, D.968b no. 1 in C major (op. 121)
there is a marked difference between the use of decresc. and dimin. In the March itself
Schubert uses only decresc. in order to not loose drive and energy. The only dim. indication is
found at the end of the delicate Trio, before returning to the March.
Schubert's songs are full of both dim. and decresc. markings. Without an exception they imply
the same interpretation. In Ellen's Zweiter Gesang, D. 838, for instance, the dim. causes
hunting horns to die away a little before the voice comes in with "Jäger, ruhe von der Jagd!"
(at the end of the 6 bar introduction). The horns are then heard from a distance in m. 10,
marked ppp (implying the use of a moderator on Schubert's piano).
63
Schubert. Ellen's Zweiter Gesang, D. 838, mm. 1-10
Franz Schubert's Werke, Serie XX: Sämtliche einstimmige Lieder und Gesänge, No.472
Leipzig: Breitkopf & Härtel, 1894-95 (Mandyczewski).
The same happens before the second verse ("Schlaf', der Hirsch ruht in der Höhle"). But in the
middle of the second verse on "Schlaf', nicht qual' es deine Seele" where there is no structural
reason to loose energy, the indication is decresc. The end of the song has again dim.
Schubert. Ellen's Zweiter Gesang, mm. 28-36
Schubert. Ellen's Zweiter Gesang, mm. 43-46
In every one of his chamber music works Schubert uses decresc. and dim. in the same way.
Schubert's use is very subtle and often additional markings shed light on the amount of
slowing down that Schubert wants. In the first movement of the Piano trio Bb major, D. 898
two structural places are thus marked differently. In m. 93 the decresc. keeps the pace intact.
The hesitation of m. 98, leading to a rhetorical pause, is illustrated by a dim. But in m. 192 the
transition back to the opening theme of m. 193 (not yet the recapitulation) is marked with a
ritardando. This powerful expressive tool is clearly used to make the fake recap convincing;
some bars later (m. 203) Schubert does it again at the fake recap of m. 204. But at the real
recap in m. 212 there is no ritardando.
64
Piano trio Bb major, D. 898, i/ allegro moderato/mm 90-101
Franz Schubert's Werke, Serie VII, No.2 Leipzig: Breitkopf & Härtel, 1886 (Brüll).
New York: Dover Publications, 1973.
Piano trio Bb major, D. 898, i/ allegro moderato/mm 188-193
Franz Schubert's Werke, Serie VII, No.2 Leipzig: Breitkopf & Härtel, 1886 (Brüll).
New York: Dover Publications, 1973.
Right before he coda of the last movement, the diminuendo is written out. This makes it
obvious that there is no local effect intended and it ensures a great contrast between the ppp
which finishes the phrase and the presto, fortissimo coda.
65
Piano trio Bb major, D. 898, iv/allegro vivace
Franz Schubert's Werke, Serie VII, No.2 Leipzig: Breitkopf & Härtel, 1886 (Brüll).
New York: Dover Publications, 1973.
In the Trout Quintet, D. 667, as usual diminuendos are in structural places, at the end of a
phrase of section. All decrescendos are in the middle of a phrase or at the end of a phrase that
does not need to loose energy. But in the second movement there is one exceptional passage.
Rather than deciding that this could be a misprint, the actual interpretation of the passage
seems to corroborate with the other usage by Schubert of the dim. and decresc. markings. The
dim. finisheds off the musical thought on the prolonged dominant and makes this thought
come to a conclusion on the downbeat of m. 51. The decresc. signifies the transition to the
coda phrase starting at the pp dolce, which can have very low energy and may be a bit slower
(a slowness caused by the dim.) than the phrase before. This means that although very little
slowing down is required under the diminuendo, it may still bring the movement of the piece
to a low point from where the coda creates stillness. The decresc. does not change the tempo
which has been reached at m. 51.
Schubert. Piano Quintet in A major, D.667 "Trout" Quintet. Andante, mm. 49-53
Franz Schubert's Werke, Serie VII, No.1 Leipzig: Breitkopf & Härtel, 1886 (Brüll). Plate F.S. 48.
New York: Dover Publications, 1973.
Naturally, one wonders how other composers may have used decrescendo and diminuendo. A
quick scan learns that there is no reason to believe that any other important composer used
these markings in the same codified way that Schubert did.
Beethoven used dim. in his early works and decresc. in his later works. In the few pieces in
the middle where he used both markings there is no clear difference, like in the Sonata op. 31
no 1 in G major. Beethoven uses dim. several times but only once, near the end, decresc. It is
not impossible that each of the dim. indications, all in structural places, could mean to loose
some energy and that the decresc. has the intention to make sure that there is no loss of energy
in the final phrase, closure following from the repeat of the figuration. But his choice to use
one or the other indication exclusively in his early and in his late years makes it unlikely that
he intended these markings to have a different effect.
66
Beethoven, Sonata op 31 no. 1 in G major. Second mvt Adagio grazioso, mm. 37-44.
First ed, Vienna: Simrock. n.d. (1803)
Beethoven, Sonata op 31 no. 1 in G major. Second mvt Adagio grazioso, mm. 53-68.
First ed, Vienna: Simrock. n.d. (1803)
Beethoven, Sonata op 31 no. 1 in G major. Second mvt Adagio grazioso, mm. 92-98.
First ed, Vienna: Simrock. n.d. (1803)
Beethoven, Sonata op 31 no. 1 in G major. Second mvt Adagio grazioso, mm. 113-120.
First ed, Vienna: Simrock. n.d. (1803)
67
V. The question of subito dynamics.
1. The fortepiano's natural decay
It is perhaps a surprising feature of the classical works before 1800 that there are very few
indications for bringing the volume down. Wy is that so? One of the most striking features of
the early piano is the rapid decay of the tone, which all by itself is one of the main defining
features of the instrument and its style. The classical phrase, moreover, has a lot of reason for
decay. Like in speech, where practically every word has a soft ending, musical motives do not
usually end with a loud note. In fact, most musical motives have a dynamical highpoint near
or in the middle. Finally, the musical phrase itself has a natural decay: the climax of the
classical phrase falls at ca 3/4 of the length of the phrase. In a regular 8 bar phrase, the climax
is likely to fall on the downbeat of bar 7 (exceptions abound, obviously: composers know this
feature of musical diction all too well). This means that there are 0two bars left for the tension
to go down at the end of the phrase, naturally including the volume to go down. In slow
movements the high point may fall in bar 6, leaving more space for winding down.
2. Beethoven
Beethoven's piano works are replete with passages in which a f or ff is followed within the
space of one, two or more bars without additional dynamic marking, by a p or pp. Modern
performance practice has it that these moments are 'subito pianos'. Based on this assumption a
style of performing Beethoven has developed which does not take the other possibility,
namely that a decay is implied, into account. But this other possibility is musically very
attractive: the f or ff may indicate a climax. To let go after a climax means that the climax is
defined better (if not, the result is just loud playing) and to regard musical gestures as musical
words make the music more speech-like, i.e. rethorical. Similarly, when a crescendo is written
before a piano passage and to modern performers there seems to be no doubt about the subito
nature of the next p, the rinforzando feature of the crescendo, described above, may be called
for and p may be reached with a natural decay.
Beethoven's Piano Trio in C minor, op 1/3 is a brilliant and often spectacular piece. The
notated dynamics range from ff to pp, and the work is full of sf, fp and rinsf. markings,
highlighting the dramatic content. Yet, there is not a single crescendo or diminuendo marking
in the entire piano part. Instead, Beethoven used an occasional hairpin to indicate large messa
di voces (3 times: i/206-212 and 317-318; iii/82-84); two crescendo hairpins (iv/234-238 and
iv/306-310) and only one (!) diminuendo hairpin (iv/230-233). In all, there are far fewer
hairpins in this trio than in the much milder first trio in Eb major and the light hearted second
trio in G major. It is of course unrealistic to believe that in this magnificent, theatrical work
exploding of expressivity, Beethoven did not ask for more dynamical development than these
few cases. Nevertheless, in modern performance practice the notated dynamics are all too
often treated in an absolute way.
68
Beethoven. Piano Trio in C minor op 1/3 (1793-94). First movement, mm. 348-353.
Henle Urtext (1969).
In the coda of the first movement of in C minor Piano Trio the ff outburst in all three
instruments (m. 350) is followed by a return of the pp which had been prevalent before the
fortissimo. No diminuendo is indicated - at first glance. But the harmony is a dominant-tonic
release (on a pedal point); besides that, both cello and piano right hand have a slur, meaning a
decay. The right hand scale descends, implying diminuendo; and the left hand chord has its
natural decay, which makes is less logical for the right hand to try to maintain its ff. The cello
is marked cresc. leading to the notated f outburst, clearly to counter the natural tendency to
create a messa di voce in each bar of this figuration (by the way, the cello part in this first
movement contains the only crescendo markings in the entire set of three trios). After creating
the climactic downbeat of m. 350 also the cello will want to diminish in order to make sense
out of the f marking - or else the question comes up why Beethoven did not notate f at the end
of the bar, if the crescendo should continue? So, although technically speaking the fortissimo
could be maintained in the right hand and a subito p could be achieved, Beethoven did not
have to further specify his intentions to the contrary as the other indicators are strong enough.
Beethoven. Piano Trio in C minor op 1/3 (1793-94). First movement, mm. 310-315.
Henle Urtext (1969).
A soft landing on I6 in m. 313 occurs after four loud bars of broken octaves. The keyboard
reaches an ff climax on the downbeat of m. 311 and since it is the highpoint, the next two bars
must go down (albeit not necessarily all the way to the p level). The sf in the strings on the
downbeat, followed by a slur in the strings in m. 312 is helpful to prepare the lighter ending
before the p in m. 313.
In the second movement of Beethoven's Frühlings sonata op. 24 for violin and piano there are
a number of puzzling crescendo markings.
These crescendos could all be interpreted as continuing to the ending of the bar, as indeed
sometimes is done in modern performance. However, on second thought there may be
69
grounds on which to decide differently; all of these grounds have been described in the
paragraphs above.
Beethoven. Sonata in F, op 24 Frühling (1800-01). Second movement, mm. 5-8.
Henle Urtext (1978).
The crescendo in m. 6 is a local crescendo, meant to reinforce the ascending interval Eb-G in
the melody, and functions therefore as a local rinforzando. The descending florish on the third
beat releases the tension and diminishes to p. Several older, famous interpretations of this
passage agree with this interpretation (see the listings below).60 More recent performances
seem to aim for a more spectacular subito interpretation. It seems that this style of interpreting
Beethovens crescendos in an absolute way is rather recent, belonging to a perspective in
which a composers' notation is interpreted as an instruction, rather than a description. It has
resulted in different aesthetics which put more value on contrasts.
diminuendo on the third beat of m. 28:David Oistrakh/Lev Oborin, Henryk Szeryng/Arthur Rubinstein, Gidon
Kremer/Martha Argerich
Crescendo to the barline of m 28 and sub. p on 29: Sebastian Müller/Nico de Villiers, Dora Bratchkova/Rudolf
Meister, Anne Sophie Mutter/Lambert Orkis
Beethoven. Sonata in F, op 24 Frühling (1800-01). Second movement, mm. 24-29.
Henle Urtext (1978).
The crescendo in the structurally important m. 28 and p in m. 29 (return of the opening) can
be interpreted in two ways, at least from the perspective of the pianist. The V7 harmony from
the second beat in m. 28 creates tension, which is released on the downbeat of m. 29.
However, the converging movement of the hands and the lightness asked for by the dots61
allow for another interpretation, namely a rinforzando towards the second beat where V7
appears, and a gradual lessening of tension in the remaining two beats. The long slur in the
violin, in itself a prime example of a dimininuendo, is marked crescendo as well; following
the harmony in the piano part, this could well be interpreted as a messa di voce.
Is this possible decay before reaching the next p marking also necissarily a dim. to p? It is
obvious that the character change on the new phrase does warrant a new dynamic, and not just
a dynamic that has been reached organically. This means that may still be a minimal dynamic
break. But at the same time the crescendo gesture can still get its natural shape with a tapering
60
61
All performances available on YouTube.
Until deep into the classical period, dots essentially indicate lightness. This is the topic of another lecture.
70
off at the end. Emphasizing the new character by keeping the crescendo growing until the
barline is, however, the other extreme.
Interestingly, some performances seem to make the dynamical break from f to subito p the
focus of the passage. The act of introducing sub p becomes a performer's moment. It is rather
detached from the logical rhetorical progression, as if the subito dynamic is what the passage
'is about'. This style, which in itself may be spectacular but also rather jerky, could in my
opinion not have existed in a time when rhetorics were the predominant aesthetic.
dim at the end of m. 28: David Oistrakh/Lev Oborin
sub p (mildly): Henryk Szeryng/Arthur Rubinstein, Gidon Kremer/Martha Argerich, Dora Bratchkova/Rudolf
Meister / sub p (very much): Sebastian Müller/Nico de Villiers; Anne Sophie Mutter/Lambert Orkis
Beethoven. Sonata in F, op 24 Frühling (1800-01). Second movement, mm. 44-51.
Henle Urtext (1978).
The crescendo under the two bar violin slur in mm. 46-47 may, like in the previous example,
be interpreted as a messa di voce focussing on the downbeat of m. 47. Arguments for this
view are the slur itself, but moreover the fact that the last beat of m. 47 is a harmonic release.
The diminuendo connected to this release does not necessarily bring the volume down
completely to p; like in the example above some break is still possible.
With dim on the last beat of m. 47: David Oistrakh/Lev Oborin, Henryk Szeryng/Arthur Rubinstein
sub p on m. 68: Gidon Kremer/Martha Argerich, Sebastian Müller/Nico de Villiers
very much sub p on m. 68: Sebastian Müller/Nico de Villiers, Anne Sophie Mutter/Lambert Orkis
Beethoven. Sonata in F, op 24 Frühling (1800-01). Second movement, mm. 62-69.
Henle Urtext (1978).
71
Each of the slurs in mm. 65, 67 and 69 may be seen as a messa di voce, with a rinforzandotype crescendo in the beginning to emphasize its expressivity. Arguments for this view are
first of all that the normal dominant-tonic release calls for a diminuendo, as every treatise or
modern textbook on interpretation teaches, and in fact every sensible musician practices.
Besides the slurs implying a diminuendo for the end of the slur, this is the coda, so creating
new tension is not in order.
Measures 65-67-69 with dim on the third beat: David Oistrakh/ Lev Oborin, Henryk Szeryng/Arthur Rubinstein
all figures sub p: Gidon Kremer/Martha Argerich, Sebastian Müller/Nico de Villiers, Anne Sophie
Mutter/Lambert Orkis (m. 65 very much)
all sub p very much: Sebastian Müller/Nico de Villiers
3. Actual subito piano
Subito piano does exist, of course. It just does not appear as often as practiced in modern
performance. For a subito piano, meaning to break normal diction and to stop the flow of the
speech, one or more out of three requirements need to be met:
-a harmonic break
-a melodic break
-a rhythmical break
In the previous chapter it was shown how dots after a crescendo marking leave little doubt
about the subito quality of the following piano marking (Beethoven's An die Ferne Geliebte
op 98. III: Leichte Segler in den Höhen). But what to do when there are no dots?
In the fourth movement of Beethoven's C minor trio op 1/3, the crescendo from m. 234 to m.
237 is followed sharply by a pp in m. 238. There is no doubt about Beethoven's intentions; the
logic is in the melodic break (subito back to thematic material) and the re-entry of rhythm, i.e.
the rhythmical break.
Beethoven. Piano Trio in C minor op 1/3 (1793-94). Fourth movement, mm. 233-243.
Henle Urtext (1969).
72
In the Waldstein Sonata op 53, the coda starts pp (m. 261, in this edition line 2, m. 3) but is
introduced in the previous bar by a repeated dominant octave, marked crescendo. The tension
goes up inside the dominant (turning into a dimished chord in m. m. 260): there is no way the
volume can go down so the pp must be subito. The result of this break in melody and rhythm
is that the pp is charged with the dominant tension and the drive of the previous repeated
octaves.
Beethoven. Sonata in C major op 53 (1803-04). First movement, mm. 254-264.
First edition: Bureau des Arts et d'Industrie, n.d.(1805).
Subito fortes and subito pianos interrupt the normal flow of the musical diction: this
disruption is the main purpose for the subito dynamic. One such interruptio is found in
Beethoven's Rondo op 51/2 in G, where the gesture of mm. 105-106 does not come to its
conclusion but instead is interrupted by a literal repeat of mm. 107-108. After repeating, the
gesture does reach a climax in m. 109. Because of the necessity to interrupt the flow, m. 107
will be subito p.
Beethoven. Rondo in G, op 5 no. 2 (1797). mm. 105-113
First edition. Vienna: Artaria (n.d.)
A specific case of subito piano is when a motive or phrase is repeated as an echo. Leopold
Mozart mentioned 'Light and Shade' (see the quote on p. 18). One such case is found in the D
major Flute Quartet by CPE Bach, first movement, m. 44, where a forte motive is literally
repeated in piano.
4. Subito fortissimo
Just like subito piano exists, subito forte also has its place. While crescendoing is an important
element of the romantic tradition in which musicians today have grown up and therefore
making crescendos, even at the end of bars, is more accepted, it seems that subito outbursts
are only possible when meticulously prescribed by the composer. One case suffices to show,
however, how even here a performer can easily go astray.
73
Beethoven. Piano Trio in C minor op 1/3 (1793-94). First movement, mm. 191-204.
Henle Urtext (1969).
Seen from the piano part, the repeated left hand G in mm. 197-199 of the first movement of
the C minor piano trio op 1/3 could easily stay p for three bars so that the ff outburst in m. 200
is a great surprise. However, the cello part has a crescendo marking, a remarkable one (see the
discussion above): it can not be interpreted as a messa di voce because of the dots following
the cresc. marking. The ff chord in m. 201 is a I6/4 harmony, releasing to a forte V in m. 202.
This V-I6/4-V movement is prepared in the cello part together with the pedal point in the
piano, together creating tension. It is not musically logical to have the cello prepare this
outburst all by itself: the repeated dominant in the piano left hand creates an undeniable
tension. The logical conclusion is that also the left hand must have a crescendo and help
prepare the fortissimo.
74
VI. Case Studies
It is time now to take the test. Is it indeed true that dynamics are often implied, rather than
written out, and if so, how do we know which dynamics to apply where?
Johann Joachim Quantz (1697-1773) wrote in his Versuch einer Anweisung die Flöte
Traversiere zu spielen (1752) about the responsibility of the performer for applying the
correct dynamics.
If you learn to recognize the different effects of the dissonances; if you pay
close attention to the repetition of the ideas, to the held notes that relieve
the flow of quick notes, to the deceptive passages that appear so often at
the cadences, and to the notes raised by sharps or naturals or lowered by
flats that lead to a foreign tonality: I am convinced, I say, that you will then
easily be able to divine when to use the Piano, Mezzo Forte, Forte, and
Fortissimo without their being written out.62
1. Mozart Sonata K. 283, Adagio
The forte indication in the opening of Mozart’s Sonata in E-flat, KV 282, is nothing more
than a general character indication. The dissonance in the first beat of measure 2, the seufzers
(sighs) in every measure, the harmonic tension and its (though sometimes temporary)
resolution, results in continually changing, subtle dynamic contours that build to a single arc
of tension over these three measures.
Mozart. Sonata in Eb, K. 282 (1775). First movement, mm. 1-3
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
The delicate opening phrase has no dynamical indications after the initial opening forte.63
This forte has the implication of 'powerful expressivity' rather than loudness in the sense of
decibels, since the notes themselves are not notes that can be performed loudly on a
fortepiano. The forte therefore stands for the expressive tension of the phrase: its high register,
lack of bass notes, thin texture, refined horizontal voice leading, and adagio expressivity.
Measure 1.
The opening slur is a seufzer as described by Leopold Mozart: heavy on the downbeat, light
on the resolving note. Although the second note C'' is not a harmonic resolution, it resolves
the tension of the long downbeat: the C comes late (a gesture reinforced by Mozart's dotted
rhythm) and is softer.
62
Quantz. On Playing the Flute. Ch. XVII section VI: 'Of the Duties of those who accompany a concertante part'.
§14, p. 228.
63
This is an original marking, according to the NMA Critical Report.
75
-The second beat Bb can be treated in tow different ways. It could be seen as a syncopation, in
which case it will receive an accent. Alternatively, it can be treated as a repeat within the tonic,
and then this Bb will receive less stress.
-The slur from Bb to Eb must be diminuendo. This means that the higher Eb is performed
with a delicato expressivity rather than a romantic and louder expressivity.
-The last note of m. 1 has a wedge which is an indication of lightness, rather than the modern
interpretation of staccato, or even staccatissimo.64 Although this note is dissonant, it needs no
extra stress as the lightest note of the bar. This lets the first bar get lighter as well as softer
towards the barline (while keeping direction to the first dissonance in m. 2).
Measure 2.
-The downbeat of m. 2 is the first dissonance on an important beat. The notation is typical for
the period: the appoggiatura indicates added expressivity, perhaps even causing a rhythmical
adaptation of the four note gesture of the first beat, as the short slur on the appoggiatura and
its resolution follows Leopold Mozart description of a longer and louder first note and a later
and lighter second note. Since the four note gesture is all about resolving the appoggiatura, it
is possible or even desirable to extend the slur to all four notes of beat 1.
-Beat 2 ends the first part of the phrase on the dominant. But as it is also a consonant repeat of
the dominant harmony and its dissonant melody note on beat 1 it will be softer than the first
beat.
-The third beat is a harsh dissonance (in this context): it is an expressive seufzer. It needs a
louder and more expressive start. The dissonant E natural in the left hand needs, as stated by
Quantz, to be played a little bit louder as well, which adds to the expressivity of the right hand
dissonance. NB: there is no new bass, indicating the flowing nature of this seufzer and its
needs to continue. The resolution to C'' is a resolution to a temporary dominant: the tension
goes up.
-Beat 4: has another dominant and is therefore accented. The slur in the right hand
(functioning as a seufzer )is light, being the second half of the fourth beat, but with direction
towards the important seufzer on the downbeat of the third bar.
Measure 3.
-The second Bb must be treated as a syncopation and is therefore accented. Beat 2 is a tonic
but being I6 is gives no real resolution, leading to the third beat I6/4, the harmony with the
greatest tension in the entire phrase.
-Measure 3 ends with the I6/4 to V7 resolution (beat 2-3) and the even better release to the
tonic in m. 4 (indicated p). Therefore the three bar phrase ends with a diminuendo to p.
Mozart. Sonata in Eb, K. 282 (1775). First movement, mm. 1-3
Urtext der Neue Mozart Ausgabe, Bärenreiter (1986).
Measure 4.
64
This is the subject of another study. Modern interpretation has it that wedges indicate shortness (staccato,
staccatissimo) or even a sharp accent. However, the first treatise that mentions this as a possibility is from 1819
(Friedrich Starke. Pianoforte-Schule in III Abtheilungen. Mit Verbindung einer leichten Anweisung das
Pianoforte rein zu stimmen, nebst Modulations-Regeln, und einer kurzen Sing-Méthode. Opus 108. Vienna,
1819). In 1789, Daniel Gottlieb Türk still rejects the possibility to interpret the wedge as a shorter note.
76
-The first beat of m. 4 being a release and the second beat a lift of the same harmony and
therefore even lighter, the third beat forte is a real subito. There is a wonderful side to the fact
that the left hand f comes before the right hand f: if the left hand third beat is not attacked
vehemently but rather sweepingly, it pushes the right hand into an expressive forte aided by
the Schleiffer before the high note. The left hand arpeggio then seems to extend the Schleiffer
upbeat to G''. After the f, both hand have slurs and extend the essentially tonic harmony. Put
into practice, the third beat left hand arpeggio has a crescendo. There is no new tension and
the bar lightens up to the fourth beat with a diminuendo, making the p arrival on the downbeat
of bar 5 entirely organic and all but subito. (idem mm. 5 and 6).
2. Beethoven Sonata in D minor, op 31/2
The coda of the first movement of Beethoven's Tempest Sonata starts with three repeated DD
unisons, marked cresc (mm. 217-218). The following bar is pp. As there is no slur over de
unisons there is no need for a diminuendo right before m. 219 The pp may therefore be
interpreted as a subito p.
Beethoven. Sonata in D min, op 31/2, first movement (1801/02). mm. 214-228.
First edition: Simrock, n.d. [1802.
The opening phrase of the second movement has its climax in m. 7. The build up of tension
through bars 1-6 culminates in a sf in m. 6, but the highpoint on the downbeat of m. 7 is
marked p. There is no doubt about the subito piano interpretation. The p is made believable
through the rising tension carrying over into the sub p downbeat and turning it into an
internalized climax.
Beethoven. Sonata in D min, op 31/2, second movement (1801/02). mm. 1-18
First edition: Simrock, n.d. [1802].
77
The next phrase, mm. 9-18 is an extended phrase, made possible by the deceptive cadence in
m. 15 (second line, m. 6). Measure 15, the seventh bar of the second phrase, is the highpoint
also in this phrase. Although the deceptive candence does not have a dynamic, the preparation
makes it logical to play another subito p: the insistence on V7 in mm. 11-12-13-14 makes the
tension go up, like in the first phrase. But the nature of vi does not allow for a loud climax and
besides, m. 15 wants to echo m. 7. The harmonic progression after vi is such that there can
only be a diminuendo on the remaining tree bars of the (extended) phrase. The forte on the
second beat of m. 15, where G'' functions as an appogiatura to the first beat of the next bar F'',
must therefore be interpreted as an accent for just that note. This puts the third beat of m. 15
in the same p character as the deceptive chord itself and leaves room for the appogiatura G'' to
stand out and resolve in the next bar.
The diminished chord on the fourth beat of m. 26 (line 2, bar 3) receives all the tension of the
crescendo before. This means that m. 27 is sub p.
Beethoven. Sonata in D min, op 31/2, second movement (1801/02). mm. 19-29
First edition: Simrock, n.d. [1802].
However, the question is whether the very last chord of bar 26 (line 2, bar 2), also the last
note on the slur, must be the loudest note. Czerny has something noteworthy to say about this
when he explains how, inside a larger dynamic movement, 'smaller expressive signs become
part of the general increasing or dimishing of the tone'.65
Czerny Pianoforte-Schule (1839)
Underneath the example he writes "the small signs <> must increase and diminish in the same
degree as the general cresc. and dim.
65
Czerny Pianoforte-Schule. Ch. 1 Von der Anwendung des crescendo und diminuendo, §7, p. 14.
78
Similarly, in the example from Beethoven's Tempest Sonata the right hand slurs at the end of
mm. 24-25-26 have a small messa di voce: the middle note is a beat, the others are not. And
although each next one of these three slurs is more than the previous, the shape of the slur
stays in tact and therefore the crescendo does not come to its climax on the last note of m. 26
but on the note before (the fourth beat).
The messa di voce markings in mm. 27-28 refer more to the rhythmical figure in the left hand,
as the right hand keeps its shape. Measure 29 is the highpoint of the little development from
m. 27 because of the insistence of the dominant in these bars.
Beethoven. Sonata in D min, op 31/2, second movement (1801/02). mm. 30-36
First edition: Simrock, n.d. [1802].
After the downbeat in m. 29 a dim. starts. An expressive rinforzando-type crescendo in m. 30
leads to V7 on beat 2, after which a resolution to the tonic comes. The harmonic relaxation is
aided by the dim. in the left hand at the end of the slur and by the lightness in the right hand,
where the ascending and diminuendo scale figure is marked with leggiero dots. The p return
of the opening motive in m. 31 is therefore not a subito. Likewise, the crescendo on the
upbeat before m. 35 should be treated as a rinforzando for the beginning of the scale, powered
by the messa di voce on the left hand slur in m. 34 but because of its lightness and its upbeat
nature essentially diminuendo.
79
Beethoven. Sonata in D min, op 31/2, second movement (1801/02). mm. 37-62
First edition: Simrock, n.d. [1802].
Harmonically, the long slur on m. 37 is a I6/4-V resolution which means that there can be no
crescendo to the barline. The p in m. 38 cannot be subito. The crescendo in m. 48 (line 2, last
bar) prepares the sub p in m. 49; it is the same material as in the opening phrase of the
movement, but this time the crescendo is added. Crescendo in m. 49 brings the phrase to its
point of arrival in m. 50 (line 3, bar 2). This is a crescendo which goes over the barline. This
has been made possible by Beethoven by making the last slur of m. 49 portato, rather than
legato (as in that case a dim. at the end of the slur would have been in order).
3. Schubert Grazer Fantasie in C major, D. 605a
As we have seen, Schubert's use of the dim. marking is almost completely codified. Because
of its specific meaning he was very careful with the use of it. One example from the Grazer
Fantasie in C major, D. 605a (mm. 157-59) is already shown above but this Fantasy is
particularly rich in diminuendo markings.66
A new phrase starting in m. 28 is prepared in mm. 26-27 by way of a dim. The fermata is the
lowpoint of the 'letting go' initiated by the diminuendo.
Schubert. Grazer Fantasy, mm. 26-29
In m. 34 the dim. is followed by an a tempo in m. 36, a confirmation of the slowing down
caused by the dim.
Schubert. Grazer Fantasy, mm. 33-36
Much like the crescendo can have a local function (and turns into a rinforzando) also dim. can
have a local function when the phrase does not come to an end.
Schubert. Grazer Fantasy, mm. 51-54
66
All examples from the Grazer Fantasy from Henle Urtext.
80
When decresc. appears all by itself, slowing down is not in order. The rhetorical question of m.
76 should not just keep its pace but in fact push on slightly, in order to make the rhetorical
pause at the end of the bar slightly longer.
mm. 72-77
The following passage is carefully balanced with decresc. in m. 235 and dim. in 243; but the
ritardando in m. 240 shows that the loosing of energy implied by a diminuendo is not as
strong as the slowing down required by ritardando.
mm. 235-243
The expressivity of the seufzers in m. 256 is reinforced by a diminuendo. In the process, the
hemiola implied by the harmony (from the middle of m. 255 until the end of m. 256) can be
brought out in a delicate way, before returning to the main motive of this section in m. 257.67
67
Perhaps in this hemiola the last half note is subdivided into two quarters, as the two chords do not have a vii-I
relationship but are in fact a cadence vi-V to I in 257.
81
mm. 255-258
Towards the end of the Fantasy the last brilliant passage is carefully crafted to loose its agitato
energy before returning to the opening motive (the coda, starting right after this passage).
Typically, Schubert uses both decresc. and dim. to achieve his goal.
mm. 278-282
Finally in the coda, Schubert's subtle energy management is demonstrated when after his
standard decresc. and dim. he asks for a morendo, intensifying the effect of the diminuendo.
mm. 283-306
82
VII. Conclusion
While initially the application of dynamics was the domain of the performer and therefore the
notation of dynamics was sparse and general, towards 1800 composers started to write more
refined dynamics. However, even when more and more different signs came into use, the way
composers notated them was still based on earlier practice. In many passages dynamics are
not notated but implied. In these cases the perfomer must base decisions about the realisation
of dynamics on an understanding of the musical language: retorical content; the character;
harmony; harmonic and melodic dissonances; the correct performance of the slur; in keyboard
music the way composers used the register differences of the early fortepiano; the notated
touch. On the other hand, dynamics that are notated are in need of interpretation: they may
indicate the character; tutti or solo passages; or a contrast. In each of these cases the notated
'dynamic' marking is not primarily an indication of volume. Furthermore dynamics can have,
besides an effect on the volume, an additional meaning of rubato; indicate a climax; create an
unusual metrical effect; a messa di voce; mean rinforzando; imply a ritardando or make sure
that no ritardando is used.
In modern performance practice classical and early romantic dynamics are often interpreted
too literally. This has implications for the characterization, the flexibility of dynamics and
rhythm, and results in an excess number of subito dynamical changes and an overly strict
metrical execution. Performance on an early piano demonstrates how high notes in many
cases need to be treated delicately, while on the modern piano they can and often will be
played louder, in a (later) romantic spirit. The correct performance of classical slurs,
meticulously described in classical as well as early romantic treatises, relies heavily on
dynamics as well as rubato. This is an essential element of the retorical style. Modern
performance treats slurs primarily as a technical indication (legato) and in doing so often
replaces retorical classical 'speaking' by romantic singing.
Bringing into play all of the principles that govern the diction of music such as character,
touch, harmony, articulation, texture, rubato, register, can fill in necessary details in the
realization of refined dynamics in classical and early romantic works. This will affect the
rhetoric, rhythm, movement and ultimately the expressivity of the piece and will lead to a
more colourful and delicate rendition of the musical story.
The Hague, July 2014
83
Bibliography
Johann Joachim Quantz. Versuch einer Anweisung die Flöte Traversiere zu spielen. Berlin,
1752/Facs. Rep. Wiesbaden: Breitkopf & Härtel, 1988.
Johann Joachim Quantz. On Playing the Flute. Translation of Versuch einer Anweisung die
Flöte Traversiere zu spielen. Berlin, 1752/Translated by Edward R. Reilly. New York,
Schirmer, 1985.
Carl Philipp Emanuel Bach. Versuch über die Wahre Art das Klavier zu spielen. 2 vols. Berlin,
1753 and 1762/Facs. Rep. Wiesbaden: Breitkopf & Härtel, 1986.
Carl Philipp Emanuel Bach. Essay on the True Art of playing Keyboard Instruments.
/Translation of Versuch über die Wahre Art das Klavier zu spielen. 2 vols. Berlin, 1753 and
1762/Translated by William J. Mitchell (2 vols. in 1). New York: W. W. Norton, 1949.
Leopold Mozart. Versuch einer gründlichen Violinschule. Salzburg: 1756/Zweyte vermehrte
Auflage. Augsburg, 1770.
Leopold Mozart. Versuch einer gründlichen Violinschule. Salzburg: 1756/Dritte vermehrte
Auflage. Augsburg, 1787.
Leopold Mozart. A Treatise on the Fundamental Principles of Violin Playing /Translation of
Versuch einer gründlichen Violinschule. Salzburg, 1756/Translated by Editha Knocker
1948/reprint of 2nd ed. Early Music Series no. 6. Oxford: Oxford University Press, 1986.
Johann Friedrich Reichardt. Über die pflichten des Ripien-Violinisten. Berlin and Leipzig,
1776.
Johann Adam Hiller. Anweisung zum musikalisch-zierlichen Gesange. Leipzig, 1780.
Johann Peter Milchmeyer. Die Wahre Art das Pianoforte zu spielen. Dresden, 1797.
Daniel Gottlob Türk. Klavierschule. Leipzig, 1789/Facsimile Reprint. Kassel: Bärenreiter,
1972.
Carl Czerny. Vollständige theoretisch-practische Pianoforte-Schule, von dem ersten Anfange
bis zur höchsten Ausbildung fortschreitend. In 4 Theilen. op. 500. Part III: “Von dem
Vortrage.” Vienna: A. Diabelli, 1839/Facs. Repr. of 2nd ed. Vienna, 1846 / Wiesbaden:
Breitkopf und Härtel, 1991.
Sandra P. Rosenblum. Performance Practices in Classic Piano Music. Their Principles and
Applications. Bloomington: Indiana University Press, 1988. (esp. Ch X, p. 362-368).
Clive Brown. Classical and Romantic Performance Practice, 1750-1900. Oxford: Oxford
University Press, 1999.
Walther Dürr. "Notation and Performance: Dynamic Marks in Schubert's Manuscripts." In:
Brian Newbould (ed.). Schubert the Progressive. History, Performance Practice, Analysis.
Ashgate, 2003. pp. 37-51.
84
85