László Somfai
Bartók's Great Crescendos
Some Observations for Young Musicians
Béla Bartók's mature music is essentially non-Classical (though the large-scale
instrumental works composed between 1926 and 1937 may be said to have
Classical ambitions); the entirety of his oeuvre and his whole artistic personality are
fundamentally Romantic. On several occasions, I have tried to find the right form to
express this conviction of mine, but a number of my colleagues have disagreed,
whether they said so openly or not. The problem derives, in part, from the divergence
between aesthetic terminology and musicians' everyday talk. In the case of
composers, 'Classical' often implies a value judgment (more valuable and more
noble, more enduring than 'Romantic' art), and given the fact that Bartók himself
affirmed that his new style had arisen as a reaction to the excesses of nineteenthcentury
Romantic music, we ought to spare him the disparaging 'Romantic' epithet.
(It has to be noted, though, that after 1930, Bartók only distanced himself from
"excessive sentimentality", not Romanticism as such.) Yet, if we keep the dichotomy
of Apollonian and Dionysian attitudes in mind as we try to determine the basic traits
of Bartók's personality and music within the context of a great historic era of the past
(the classics of the twentieth century: Schoenberg, Bartók, Stravinsky), then Bartók's
powerful Romantic attributes begin to appear in a positive light.
All this is confirmed to me by the study of Bartók's piano recordings. One has
to admit that the general interpretive trends of the 1950s and '60s were scarcely
helpful to Bartók performers. True, the idiom of new music became more familiar;
mixed metres no longer presented any problems, and Bartók's scores began to
seem easier after the appearance of performances that were more and more
accomplished technically. In the process, the compositions themselves-
especially the string quartets and some of the symphonic and piano music-came
across as being purely 'classical', sometimes even 'classicising' masterpieces. Yet
Bartók's own playing is extremely personal, eventful and stirring. He deployed an
entire arsenal of Romantic piano techniques, made the musical characters larger
than life and strove for a powerful cathartic experience. He was able to create
catharsis several times, and at different levels of intensity, in the performance of
works such as the Sonata for Violin and Piano No.2, the first movement of the
Sonata for Two Pianos and Percussion, or the Second Piano Concerto. Still, even
in his shorter piano pieces, he could shape deeply moving conclusions or
penultimate moments that the printed score would not have suggested to another
performer. Through his performances of his works, Bartók made it clear that
precise reading of the score and virtuosity were worthless if not backed by a true
artistic personality, courage and imagination.
The sound documents of Bartók's piano playing have been accessible for 25
years now, and their influence on performers may be easily traced. When
contemporary Hungarian pianists perform Bartók, one can see a certain shift in
style and find interpretations that are bolder, more persuasive and more individual
than those of the past. This first manifested itself in the interpretations of Zoltán
Kocsis, which have exerted a great influence on others. Kocsis got to the bottom
of the issues when he studied and recorded Bartók's entire piano oeuvre for
Philips. He put his own creativity to the test by asking himself how Bartók might
have played those pieces that have not come down to us in his own recording. For
younger Hungarian pianists, it is no mean task to come to grips with this dual
model-Bartók's recordings and Kocsis's Bartók interpretations. In other genres,
there is only sporadic evidence that the message of Bartók's recordings have been
artistically processed; it is enough to listen to Boulez's recent Bartók CDs of the
piano concertos or the majority of recent string quartet recordings.
It seems to me that the study of Bartók's compositional methods and working
habits reveals another way in which the Romantic nature of his personality
manifests itself. How did Bartók compose? In 1925, the writer and poet Dezsô
Kosztolányi asked him, "How do you work? Systematically?" Bartók answered,
"I can say this: I don't like to do two things at once. If I begin something, then
I live only for it until it is finished." "At the desk? At the piano?" "Between the desk
and the piano."1 Or should we believe the later recollections of the musicologist
Antal Molnár? According to him, "Bartók composed like Mozart-he worked out
everything in his head and then wrote it down with fiery speed." The latter
comment receives no support at all from the documentary evidence. With my
knowledge of the entire manuscript material, I can safely say that, under normal
circumstances, Bartók began work on a new composition by improvising at the
piano, searching for materials and shaping them behind closed doors. When the
beginning of a movement or a more substantial section had emerged, he wrote
down the draft, not "with fiery speed", but rather with the precision of an
experienced musician/ethnomusicologist. In an often-quoted letter of June 1926
to his wife Ditta (written after he had sent her and their son Péter off on vacation,
so he could work on the First Piano Concerto undisturbed), we read:
The other night, I was getting ready for bed around midnight when all of a sudden
I thought of something I had to write down immediately. And that was the end of my
desire, and ability, to sleep. I spent at least another hour ruminating over what I had
written. Another time, I jumped out of bed to write something down... The problem is
that I couldn't even begin the most important thing, the piano concerto. But such things
can never be forced.
All this must have sounded extremely Romantic even in the 1920s, when many
prominent composers, especially those belonging to the Schoenberg circle,
started work only after a stage of rational planning and preparation. In those days,
the craftsman-like attitude of a Hindemith, who could write a new piece in the
dining car of an express train between Berlin and Frankfurt, passed for a virtue.
The composer himself had boasted of his feat. Bartók must have felt out of place
in such company. After all, he was not a full-time composer and could not afford
to sit at his desk every morning, all the year round, to fulfill his daily quota. His
experience must have been that finding the main idea and the opening materials
of a new work was almost always unpredictable. Then, after a period of planning
and preliminary thinking, the opening material would finally crystallise as if by
miracle and now there was something that could be further polished, improved,
made into a complete piece. Bartók himself probably did not understand how this
happened, and we must believe that he did not want to understand it. The study
of his manuscripts confirms what we know from the printed scores to be one of
the most salient aspects of his art-namely, that he begins each important work
with a new theme or character that has not existed before, strong and
emphatically original. These themes or characters were not made but rather found
by the composer who, after the necessary mental prepration, had to wait for this
to happen. As the work continues, technique may take over; personal habits of
shaping the material, tried-and-true harmonies and favourite rhythmic processes
may come into play. Those are the elements that constitute Bartók's personal
style, the totality of the fingerprints usually described and catalogued by analysts.
Of course, one can be confident about such observations only after decades
of research and repeated, systematic investigation of every single extant manuscript
of a Bartók composition. Yet, I would like to share my conviction with young Bartók
players who are all too often lured by the siren voices of analytical trends not
primarily geared towards the things that really mattered to Bartók, but which rather
emphasize the search for abstract proportions, the labelling of certain technical
fingerprints or an excessively literal identification of folk music models.
A study of Bartók's manuscripts may be helpful in understanding which
elements were crucially important to him. There are many reasons why sensitive
musicians should want to see composers' manuscripts, whether they are by Bach,
Mozart, Liszt, Brahms, Stravinsky or Bartók. In some cases they might suspect a
wrong note, or else they might want to see first-hand whether an irregular slurring
pattern is authentic. They might also be curious as to what was corrected, deleted,
or added in the course of the compositional process. And in general, they would
like to see the visual image of the score, the style of the handwriting, in order to
come closer to the composer. With regard to Bartók, the situation is rather
complex, but very exciting. Not only string players, but pianists and conductors as
well would greatly benefit from consulting Bartók's first draft for the opening
movement of his String Quartet No.4-a composition crystal-clear in its layout
and elaborated in great detail. This is a three-and-a-half page long document
written in ink. Here we see what Bartók considered the most important thematic
material and what he did not notate (did not decide on) for the time being. Then,
one should look at the next autograph draft, which runs to eight full pages, and
contains all the filling materials, transitions and developments. It will turn out that
some of the canonic imitations and other 'clever' techniques, so impressive in the
score, only serve to give form to a texture. There is even a third autograph manuscript,
the fair copy (here the first movement takes up ten pages), in an easily
readable, larger hand, where the crucial performing instructions concerning
articulation, dynamics, accents and tempo finally appear. With three autograph
scores, this is a complicated source situation. Yet even that is not the complete
picture, for Bartók made some more small, but crucial changes in the copy sent to
the publisher, and even more in two sets of galley proofs. In other words, the main
interest of the Bartók manuscripts-unlike with those of Bach, Mozart or Haydn-
does not lie in double-checking suspected wrong notes and slurs (a task that the
editors of the critical edition of Bartók's complete works will undertake), but
above all, in allowing us to approach the composition from up close.
The dynamic markings and accent signs are particularly revealing. To pick out a
single symbol, the crescendo/decrescendo hairpins appear in most printed scores as
uniform signs, which is an artificially regulated version of what Bartók had intended
for the musician to see. Example 1 shows a passage from the First String Quartet's
opening movement in the autograph score-the climax of the Lento's first section,
preceding the molto appassionato, rubato viola theme of the middle section. For the
sake of comparison, Example 2 shows the same passage as printed by Editio Musica
Budapest, to illustrate how these hairpins were treated by a professional engraver
who is familiar with the traditions, general rules and aesthetic principles of musical
notation. (NB: This revised edition contains some authentic tempo instructions
which are not in the autograph, but were added later.)
The wide hairpins in Bartók's handwriting are beautiful and suggestive: they
indicate a powerful increase in volume or an enormous-and rapid-loss of
intensity. In some cases, they are tilted up or down, following the direction of the
melody. Some are thick, others thin. (It is not generally known that in the printed
edition of the Fourteen Bagatelles, from the time of the First Quartet, Bartók
expressly asked the engraver for two kinds of crescendo/decrescendo hairpins:
thick and regular.) The two-note slurs in the "Hungarian" rubato melody for
the viola
look so much more exciting in the autograph! Any musician with a keen eye will
notice that a number of notational subtleties got lost in the printed version;
around this time (1908)
Bartók was not yet the
hawk-eyed, meticulous
proof-reader he later
became. The word molto,
written inside the giant
crescendo hairpin of the
cello, is not there in print; in
the next measure, the
espressivo melodic step of
the second violin (originally
mezzoforte) erroneously
became piano; in the first
measure of the next system,
the tenuto signs are absent
in the second violin part; the
first of the cello's open fifths
(C-G) lost its sforzato accent; several vertical marcatissimo signs were converted
into weaker, horizontal marcatos, etc.
If someone were to link these vigorous crescendo and decrescendo hairpins to
Bartók's truly 'Romantic' phase from January 1908-a style saturated with memories
of Tristan, which helped the young composer get out of his system his grief over Stefi
Geyer-I could counter with any number of wide hairpins from the manuscripts of
the later string quartets, piano and symphonic works. I could also cite one of Bartók's
strict injunctions to the printer of his Piano Sonata (1926), where he insisted that the
crescendo and decrescendo hairpins be engraved with an opening just as wide as they
were in the manuscript, and thicker than usual. For an example of Bartók's style in
the 1930s, consider the (Second) Violin Concerto. Example 3 shows an excerpt from
the first movement in the Boosey&Hawkes score-the crescendo sign accompanying
the solo violin's glissando at the end of the Calmo section with the dodecaphonic
theme. Here Bartók wrote on the
margin of the galley proofs that he
wanted a hairpin twice as wide as
before. (In the published edition the
hairpin was, in fact, modified, though
not exactly doubled in width.)
A correct musical text is one thing;
a score whose visual appearance in
itself inspires the performer is quite
another. The issue of the wide hairpins
is but one component of Bartók's
elaborate ideas about performance,
many of which, unfortunately, can
only be seen in the manuscripts. Yet
this one issue might suffice to
encourage a young artist who reads
Bartók's scores with fresh eyes to try
out some "heresies" in order to make his or her performance more expressive,
more fiery, more personal.
My partner in this imaginary conversation, the enthusiastic young musician,
will ask, "What scores should we use to play Bartók? Don't we need facsimile
editions more than anything else?" I am certainly a proponent of facsimile
editions, accompanied by the necessary commentaries.2 At the same time, given
the complex source situation of certain Bartók works, I would caution against
using the autograph (any autograph) as an exclusive authentic basis for the
performing musician-the kind of basis a Mozart or Haydn autograph would
provide. With Bartók, the text was not definitive until it was printed or revised. The
young musician might ask whether the new printed editions are better than the
old ones when it comes to doing justice to Bartók's performance instructions.
There are no easy answers to this question. First of all, even those first editions
which had been proofread by Bartók are not all alike in this respect. With regard
to the crescendo/ decrescendo hairpins, the score of the Piano Sonata by Universal
Edition is certainly better than many other Universal Edition scores (though only
a seasoned Bartók scholar would know this). Some of the string-quartet parts are
better than the full scores, which often have a crammed look. In the Universal
Edition and Boosey&Hawkes scores revised by Peter Bartók, hairpins are often
corrected so they do not stop before a barline but lead straight into the new
dynamics as Bartók had intended-yet, the width of the hairpins is usually not
taken into consideration. The Bartók complete critical edition in preparation will
widen the hairpins within reason and make them more 'Bartók-like', but a hairpin
like the one seen in the manuscript of the First String Quartet would look absurd
in print. Bartók's scoring was extremely dense; he crammed more measures in a
single system than a printed edition can accommodate in an aesthetically pleasing
manner. The hairpins in the complete edition, though relatively wide, will have to
be more elongated.
One of the inhibitions that becomes an obstacle to an artistically free, expressive
Bartók interpretation has to do with tempo. I am certain that today's young
Hungarian musicians have received much better guidance about authentic Bartók
tempi than their teachers' generation had. They know that Bartók attached special
importance to tempo, despite evidence that in some cases may seem confusing.
A well-chosen tempo is a defining element of the character of a given movement
or theme. It is natural these days to use a pocket-size digital metronome and to
take the composer's own tempo markings as a starting point when practising a
piece by Bartók. When checking a recording (our own or someone else's), the
metronome or stopwatch will come out of its case and, at the slightest discrepancy,
the verdict will arrive quickly, Beckmesser-fashion: the tempo is not
authentic; Bartók's indicated durations are not being observed. (On a subjective
note: it is easy to notice discrepancies in tempo, but much harder to gauge
dynamics, accents, intelligent articulation, character or the balance between main
and subsidiary voices. If a performance-conducted, say, by Sándor Végh or
Nikolaus Harnoncourt-is really significant and reveals the depths of Bartók's
scores, it no longer matters to me whether or not the tempo is exactly identical to
what the score says it should be.)
No doubt, we have turned the question of Bartók's tempi into a fetish. In an
earlier article and in the last chapter of my book,3 I adduced a great deal of
evidence to show the impossibility of interpreting all of Bartók's tempo markings
in the same way, since those markings come from different times in his life,
belong to different genres and appear in editions with different publication
histories. There are countless typographical errors, from the famous case of the
Adagio barbaro to any number of data that were entered wrong and never
corrected, even in the American years.4 These should be enough
to shake the confidence of every committed Bartók player who wants to believe
firmly that the
author's markings can never be questioned. Two famous metronome markings
deserve to be mentioned here in particular. For decades, the world's most famous
conductors-Hungarian and non-Hungarian alike-have tried to interpret the
second movement of Concerto for Orchestra convincingly while following the
marking Allegro scherzando, 
l
=74, which was more like allegretto. Yet Bartók's
manuscript has l =94! What happened was that in the bound photocopy, mailed
from the United States to the London offices of Boosey & Hawkes, the upper
semicircle of the digit 9 had been cut off. The number 94 became 74 and was
engraved that way. Bartók read the galley proofs, but failed to notice this fatal
error. It has to be noted that the discrepancy became apparent as soon as
Koussevitzky's recording, made shortly after the world premiere from the
handwritten score, was released on CD. Another case has been a major headache
for pianists. The balladesque third movement of Improvisations, marked Lento,
rubato, bears the metronome marking l
=96 in the printed score. Yet there is no
doubt that Bartók had written l =69. The two digits were transposed by the
engraver and the error, once again, escaped Bartók's attention. It is the only
one
of the work's eight movements of which we do not have a recording by the
composer. For a long time it was up to the pianist to decide whether they should
adopt the rather hectic rubato suggested by the print, or rather follow their
instincts and disregard the metronome marking in the score. Peter Bartók's
corrected edition gets the numbers right; the new editions supervised by the
composer's son are, therefore, a must for any performer today.
I have spoken elsewhere of the tempi heard on Bartók's recordings, which
(even the main tempi at the beginning) often significantly differ from the printed
scores. At this point, I do not wish to go into the possible causes of these
discrepancies or how today's young performer should deal with them. Yet I want
to draw attention to an important fact, and to a Bartók statement that is difficult
to find. We must know that Bartók couldn't, or wouldn't, make a final decision on
the ideal tempo until the piece had been performed; this was true even in the
genres where he could count on good performers, well versed in his music. For
instance, when he sent the engraving copy of the Third String Quartet to his
publisher in Vienna in the autumn of 1928, he stipulated that "it cannot be printed
until it has been performed and I have heard it"; on 24 April 1929, he declared,
"I wasn't able to revise the metronomic markings until now, during the rehearsals
of the Waldbauer-Kerpely Quartet; many of the markings had to be corrected." The
issue, here and in numerous other cases, was not to slow down a tempo to make
it easier to play. On the contrary, he made the Prima parte faster; the allegro in the
Seconda parte stayed the way it was, but some of the Piu mosso's were accelerated.5
The statement that is difficult to find concerns the correct interpretation of
metronomic markings and the timings, precise to the second, which appear in
Bartók's scores. This statement was published in English in the 1941 piano
reduction of the Violin Concerto. Who would look there if a full score is available
to the student and a printed part to the soloist? Yet it is very important that every
Bartók player should know it.
Timings, noted from an actual performance, are given for sections of movements and,
at the end of each movement, for the whole thereof. It is not suggested that the duration
should be exactly the same at each performance; both these and the metronomic
indications are suggested only as a guide for the executants. It appears to me better to
present them as exact timings, rather than round off the figures.
It is one of those declarations where the outlook of the composer is inseparable
from that of the ethnomusicologist, who had painstakingly notated
thousands of folksongs, constantly revising them over the years. Bartók knew that
every performance was unique, both in folk practice and in the concert hall, but
he still thought it better to document the specifics of a given performance than to
give approximations.
An expressive performance in Bartók's spirit entails more than the choice of
correct basic tempi; even more important is, perhaps, the observation of tempo
fluctuations-segments played somewhat more slowly, followed by a return of the
original tempo. For the most part, Bartók did not indicate these fluctuations as
downright tempo changes in the score; a good musician must feel these things with
the help of his or her schooling and taste. Also, these tempo differences cannot be
measured exactly; they are more like agogic emphases. Many of his recordings-
the two Allegro barbaro renditions, numerous pieces from Mikrokosmos and, of
course, the performance of Beethoven's Kreutzer Sonata-strikingly demonstrate
how Bartók took themes emerging from ostinato motion as well as from
transitional sections, slightly more slowly than their immediate context. After all,
such segments are more important, more individual. Bartók even wrote about this
when, in 1931, he answered a query from Max Rostal concerning the correct tempi
in String Quartets Nos. 1 and 4. This is
what he had to say on one particular
passage in the first movement of the
Fourth Quartet (see Ex. 4):
In measure 37 of the first movement, the
chords may, and even must, be played
much more powerfully; as a result, the
tempo will, of course, become more drawn out. Here and in similar places, it would be
confusing to indicate tempo changes; the tempo changes by itself, so to speak, if we
understand and render the character of these pesante chords correctly.
About 2,000 printed pages of Bartók's correspondence have so far been
published and at least as many are still awaiting publication. Yet in all this
material, there are no more than a dozen or two statements like this one, containing
otherwise unobtainable information from the author, addressed directly to
the performer. These should be published in the preface of every score; they will
all be included in the critical edition.6
In his above-quoted letter, Bartók cautioned Rostal regarding the performance
style of the second-movement Prestissimo, con sordino of the Fourth Quartet,
"with legato bowing, of course, definitely not spiccato!" In the postscript to a letter
dated 23 October 1934, to the violinist Rudolf Kolisch, who was then preparing the
world premiere of the Fifth Quartet, he wrote:
PS I am not very fond of spiccato playing. Any passages not marked with a @ should be
played on the string, separating the notes to some extent (or hardly separating them)
depending on the character of the passage. The note groups marked with ..... (dots)
could, perhaps, be played spiccato.7
In his memoirs, Zoltán Székely also repeatedly mentions on-the-string playing,8which might be called Bartók's 'default' string technique. If we listen to the
opening of the Second Rhapsody's Friss section in Székely's performance, of
which Bartók thought so highly, and compare it to the readings of other Hungarian
violinists whose style has been deemed "authentic", it turns out that many
otherwise great violinists who claimed to be genuine Bartók players (including
André Gertler) were not authentic witnesses at all. I add with some trepidation that
even Sándor Végh, during his years as a quartet player, occasionally forgot the
instructions he had received from Bartók, or revised them in the spirit of the new
times, particularly when it came to tenuto playing.9
The young Hungarian string player-my imaginary conversation partner-
might interject that he or she is familiar with these documents, has studied the old
recordings and attended the masterclasses of Lorand Fenyves and others. Yet we
cannot go back in time or undo all that has happened in the development of violin
technique since the 1940s, as a result of which the string instruments today sound
more intense and secure in larger concert halls, whether in solo or in quartet or
even in orchestra. Of course, a Bartók scholar could not change the tastes of
today's string players or the tastes of the audiences influenced by these players-
nor would he ever want to effect such changes. Yet it is to be regretted that a
variegated, refined and discriminating sound, or a multifaceted right-hand
technique that does not hesitate to take risks, is no longer considered a virtue.
Such features were found in the performances of the best Hungarian soloists and
chamber musicians of the early twentieth century, and they contributed to the
excellence of such Bartók interpreters as Stefi Geyer, Jelly d'Arányi, Joseph Szigeti,
Zoltán Székely and Ede Zathureczky. (We can still admire these qualities in the
cello playing of Miklós Perényi.) Many people think that audible shifts are
antiquated; yet they were an essential feature of any expressive performance, and
they were very much part of Bartók's consciousness (as well as Schoenberg's, by
the way), while continuous vibrato was practically unknown. Artists playing
Bartók today must be aware that the composer counted on a wide range of
timbres, playing techniques and nuances. They must take this into account and try
to translate it into modern technical language, while avoiding significant losses if
possible. One should acknowledge, too, that producing a consistently beautiful,
intense sound is no substitute for crystal-clear intonation, especially in a string
quartet. In the above-quoted last chapter of my book, a chapter intended for
performers, I pointed out that only instructions like espressivo, molto espress.,
ma
con calore, cantabile, etc., call for a genuine vibrato that is so ubiquitous today.
Even grazioso and dolce passages should not be played with too much vibrato, as
such passages are rather far removed from espressivo.
To focus on string-quartet playing for the moment, do we have any authentic
recordings from Bartók's lifetime, made under his supervision, that would back up
these claims? Unfortunately, we don't. It is one of the major lacunae of interpretation
history that we have no recordings whatsoever of the Waldbauer-Kerpely
Quartet, which premiered the first four Bartók quartets. The Fifth and Sixth
Quartets were premiered by the Kolisch Quartet; we only have a Kolisch recording
of the Fifth, but by the time that recording was made, the personnel of the quartet
had changed, and the new foursome had only rehearsed the Sixth Quartet with
Bartók, not the Fifth. Nor is there a recording of the Fifth Quartet with the New
Hungarian Quartet (Sándor Végh, László Halmos, Dénes Koromzay, Vilmos
Palotai), which gave it its first Hungarian performance after rehearsals with Bartók.
To be sure, some of these players later played with other partners and recorded the
works. The Hungarian String Quartet, led by Székely, included Koromzay and
Palotai; Végh was to found the Végh Quartet. There were only four string quartet
recordings made in Bartók's lifetime. These are highly recommended to young
Bartók players, though their source value is only secondary.10
Not surprisingly, there is hardly any trace of the instructions Bartók gave to
Hungarian artists; he could talk to them in person at the Academy of Music during
rehearsals.11 However, he sometimes shared important details with foreigners.
One particular warning, addressed to Kolisch in an as-yet-unpublished letter on
the Fifth Quartet, shows how important the correct performance of upbeats and
downbeats was for Bartók.
As in the first movement of the Fourth Quartet, barlines often serve merely as
orientation points; to avoid misunderstandings, I frequently use the sign; what
precedes the is an upbeat, what follows it is a downbeat (mostly accented).
I doubt whether performers of today, Hungarian or not, always read and
accentuate Bartók's scores in this way. One of the most exciting Bartók quartet
performances I have heard in recent years is the Zehetmair Quartet's (Thomas
Zehetmair, Ulf Schneider, Ruth Killius, Françoise Groben) recording of the Fourth
Quartet.12 It is very inspiring, particularly in their rendering of the special muted,
plucked, sul ponticello and glissando sounds. Yet in the fast movements they play
many measures of music practically without any accentuation at all, as a single
undifferentiated surface. It is like reading a poem as prose, disregarding verse
structure and rhymes, intentionally proceeding too fast and creating new connections
among the elements. This may be interesting and novel, but the original
is perhaps better, or at least is still valid. Even though Bartók's scores appear to
be constructed with the precision of an engineer, the notes, measures and phrases
are not all equal. A performance must breathe. If a listener feels tired too soon by
an interpretation that is too tight and has no relaxed moments, if he or she only
hears masses of sound and virtuosity instead of meaningful musical sentences,
then Bartók has fallen victim to fashion.
One day in America, Yehudi Menuhin paid a visit to Bartók to play the First
Violin Sonata for him and to ask for his advice. The composer found the
performance surprisingly good. We know the story from Menuhin's memoirs, but
the following passage from one of Bartók's letters (to Wilhelmine Creel, dated
17 December 1943) is perhaps even more revealing:
When there is a real great artist, then the composer's advice and help is not necessary,
the performer finds his way quite well, alone.
Many of us remember the First Sonata as played by Menuhin in concert in the
years after Bartók's death. Younger people may want to consult his recording or
the video of his appearance in St. Petersburg. This score was a perfect fit for
Menuhin's temperament and artistic personality; his performance was magnificent,
powerful and deeply personal. Bartók himself had played his First Sonata
with several outstanding Hungarian violinists. Jelly d'Arányi's rendition was
probably in no way inferior to Menuhin's; the composer of the sonata and its
dedicatee must have played the work more 'authentically'. Yet to Bartók, true
artistic greatness and a searing, inspired approach were more important than
'authenticity' in such details as musical 'mother-tongue' with its attendant rubatos
and agogic accents. Young Hungarian Bartók players take note: even if it is not
going to be easy to escape the long shadow of today's leading interpreters, there
is always hope.
1 
Translated by David
E. Schneider in Bartók and His World, ed. Peter Laki. Princeton: Princeton
University Press, 1995, p. 230.
2 recommend the following facsimile editions of Bartók manuscripts that I have prepared: Two
Rumanian Dances (Budapest: Editio Musica, 1974); Piano Sonata (Budapest/Vienna: Editio
Musica/Universal, 1980); Béla Bartók's Black Pocket-book: Sketches 1907-1922 (Budapest: Editio
Musica, 1987); Viola Concerto (Homosassa, Florida: Bartók Records, 1995). Furthermore, some
interesting facsimile pages may be found in one of my exhibition catalogues [In Bartók's Workshop.
Sketches, Manuscripts, Variants: the Documents of Creative Work]; Budapest: MTA Zenetudományi
Intézet, 1987), as well as in my book Béla Bartók: Composition, Concepts, and Autograph Sources
(Berkeley/Los Angeles/London: University of California Press, 1996).
3 Béla Bartók: Composition, Concepts and Autograph Sources, pp. 252-262, 279-294.
4 For instance, the duration of the Bagatelle No.2 is 1'48" according to Bartók's handwritten note-
yet there are two extant recordings with the composer which clock in at 45" and 44", respectively.
Bartók erroneously wrote 1'48" instead of 48".
5 See my "Tempo, Metronome and Timing in Bartók's Music: The Case of the Pianist-Composer",
Jean-Jacques Dünki, Anton Haefeli, Regula Rapp, eds., Der Grad der Bewegung. Tempovorstellungen
und -konzepte in Komposition und Interpretation 1900-1950 (Basler Studien zur Musik in Theorie und
Praxis 1, Bern: Peter Lang, 1998, pp. 47-71).
6 ave to point out that some of the revised editions published after Bartók's death have been
manipulated. The Rózsavölgyi edition of the First String Quartet is, in general, one of the most
problematic Bartók scores. True, it was improved by Denijs Dille in 1964, when the edition was
reissued by Editio Musica. Dille corrected many performance instructions on the basis of a list made
by Bartók in the 1930s ("Correction of Metronome Numbers, etc."), the aforementioned letter to Rostal
and a personal copy in which Bartók had made some entries. Even so, the score contains hundreds of
inaccuracies. For an example of the manipulations on the text, one should look at the first page of the
third-movement Allegro vivace. Here one finds three poco piu mosso indications, separated by two
metronome markings: ol = 88-92. These are not notations made by Bartók. All the composer had
written to Rostal was "3rd movement. From [rehearsal number] 1 to 3 somewhat faster, with the
exception of the sixth measure after 1 and the fourth measure after 2, which are more or less in the
main tempo." In other words, this is one of those cases where "the tempo changes by itself, so to
speak", as in the Fourth Quartet where "it would be confusing to indicate tempo changes."
7 Spiccato technique, which was becoming very fashionable, made Bartók see red to the point where
he even cautioned orchestral players against using it. In measure 8 in the fifth movement of Concerto
for Orchestra, where the violins begin their semiquaver runs in the Romanian style, he appended the
footnote [in English]: always non spiccato (i.e. legato).
8 Claude Kenneson, Székely and Bartók: The Story of a Friendship (Portland, Oregon: Amadeus,
1994).
9 See my "Idea, Notation, Interpretation: Written and Oral Transmission in Bartók's Works for
Strings", Studia Musicologica 37 (1996), pp. 37-49, esp. pp. 42-43.
10 The four recordings are as follows:
1925: String Quartet No.2, Amar-Hindemith Quartet (Licco Amar, Walter Caspar, Paul Hindemith,
Maurits Frank), DG Polydor 66425/28. They used the first edition which contained many wrong tempi.
Bartók only found out about this recording after the fact.
c1936: String Quartet No.2, Budapest Quartet (Josef Roisman, Alexander Schneider, István Ipolyi,
Mischa Schneider), CD=Biddulph Recordings (1995) LAB 107. Bartók never heard this recording.
c.1936: String Quartet No.1, Pro Arte Quartet (Alphonse Onnou, Laurent Halleux, Germain Prevost,
Robert Maas), HMV D.B.2379-82. Bartók valued their playing, but he didn't hear this performance prior
to the recording. It does not contain the comments regarding tempo, etc., which Bartók shared with
Rostal.
c.1941: String Quartet No.5, Kolisch Quartet (the second group formed in America: Rudolf Kolisch, Felix
Khuner, Jascha Veissi, Stefan Auber), CD = Biddulph Recordings (1995) LAB 107. Bartók knew and
valued the Kolisch Quartet in its original, European composition; this, however, is a reorganized group.
11 A couple of documents, which have come down to us by accident, attest to the fact that even
performers intimately familiar with Bartók's music occasionally had to be corrected after their first
readings. For instance, Bartók wrote out the first 142 measures of the first violin part from the Fourth
Quartet for Imre Waldbauer, maybe to test the technical difficulties of the piece. The part originally
contained only slurs and staccato markings, but in a later stage (after hearing his friend's reading),
Bartók added articulation signs-tenuto, staccato+tenuto, staccato at the end of a slur, etc.-to almost
every single note. Even Waldbauer did not at first play the "ordinary notes" the way Bartók wanted to
hear them. See my "Idea, Notation, Interpretation" (fn. 9), where musical examples are given.
12 ECM New Music Series ( 2001) 1727, 465 776-2.
László Somfai
is former Director of the Bartók Archives in Budapest and Professor of Musicology at the
Liszt Ferenc Academy of Music. His books include The Keyboard Sonatas of Joseph Haydn
(1996) and Béla Bartók: Composition, Concept, and Autograph Sources (1996).
