Spacer Zu den Kompositionen Ensembles
Unterricht Aufnahmen Presse
CHARM - Symposium sept 08: "Playing with recordings" zurÜck
History of interpretation 1913 - 1933
Beethoven’s Fifth; a passage to the 19th century

Beethoven’s Fifth – a passage to the 19th century

a conference from the 6th CHARM - Symposium - September 12

1. Ravel – Immerseel

Ladies and Gentlemen, before getting to my main subject, I would like to present you a small excursus, which I prepared for today. As you’ve certainly read, I am examining the first recordinsgs of Beethoven’s Fifth for my University in Berne in order to find significant differences concerning influence, traditions and trends in the time of early recordings, approximatly between 1910 and 1933.

My object is the history of orchestra playing and, of course, I am not just interested in this period but also in the implications for the time before the recording era. In drawing conclusions we must always bear in mind the gap between Beethoven’s time and the time which provides us with acoustic evidence of musicians habits.

To start with, I’m going to take a subject which seems to be more comfortable to talk about because there is not so much guesswork to do, since in the case of Maurice Ravel (and of course others, such as Elgar, Debussy, Richard Strauss or Holst) we do have recordings of strong authoritiy either by the composers themselves or by very close interpreters. In the case of Ravel’s Boléro we have actually two recordings, realized by two different orchestras of the same city, Paris, during the same weeks, just 14 month after the first performance. One of them is conducted by Ravel, the other by Piero Coppola under the supervision of Ravel himself.

Recording 1
Le Grand Orchestre de Gramophone – Piero Coppola
(supervised by Maurice Ravel)
13 January 1930, Salle Pleyel Paris

Recording 2
Orchestre de l’association des Concerts Lamoureux – Maurice Ravel
? January 1930

To give you an idea of my approach towards comparing interpretation I’ve chosen a recent recording which refers explicitly to the two recordings listed right at the top. Jos van Immerseel describes them in the CD-booklet as „old-fashioned“ and „largely tradition-conscious“, and his aim is to get back some of the spirit of the time with his „period-instrument“-orchestra anima eterna.

Recording 3
Anima aeterna – Jos van Immerseel
October 2005

First I compared the three recordings in detail. I then took two more productions, one realized in the same year as Ravel and Coppola (Willem Mengelberg), the other a modern (or conventional) one with Sir Simon Rattle for additional comparing.

additional recording 1
Royal Concertgebouw Orchestra – Willem Mengelberg
31 May 1930, Concertgebouw

additional recording 2
City of Birmingham Symphony Orchestra – Sir Simon Rattle

In my listening analysis I focused on the following parameters:

a. tempo and its flexibility

b. Rubato and other soloistic features (such as vibrato, glissando, articulation and dynamics)

Additionaly I tried to characterize the recordings according to formal aspects, questions of balance and differences in articulation.

a. tempo

One part of my research project is the development of a computer program, which enables me to format and track the recordings in the same way, in order to compare the various aspects in question with great ease (click&play, by Stephen Lumenta, ©2008 by Hochschule der Künster, Bern). The program also indicates the timing for each track, and in the case of the Boléro it’s the easiest thing to deduct the metronomized average tempo of each part.

Immerseel describes his approach to the tempo matter as follows: „We know that Ravel criticised both Piero Coppola and Toscanini for taking too fast a tempo for his taste in this piece. Ravel’s personal score (Bibliothèque national Paris) gives a metronome mark of crotchet=66 (whereas Durand changed the tempo to 72 in the printed score).“ (Immerseel in the booklet of his recording)

We can easily see, that Immerseel goes somewhat overboard in his quest for authenticity. His tempo is considerably slower than Ravel’s proposition and it is not as stable as Immerseel claims it to be.

Recording 1
Le Grand Orchestre de Gramophone – Piero Coppola
(supervised by Maurice Ravel) 1930
TT: 15.49 (mm between 60 and 68, average: 64.5)

Recording 2
Orchestre de l’association des Concerts Lamoureux – Maurice Ravel 1930
TT: 16.05 (mm beginning 61 - end 65, average 63.4)

Recording 3
Anima aeterna – Jos van Immerseel 2005
TT: 16.42 (mm beginning 59 - end 64, average 61.0)

You might say that the difference is marginal, but listen once to one passage and you will see how much the character of the music can change with just this tiny difference!

(Ravel, figure 9)

Of course it is not just the tempo that is so different, but most notably the rhythmical interpretation of the continuous accompanying motive, that makes the big difference! Ravel does not perform it mathematically, but always with a hint of swing and also dynamic – I would say just naturally dancing! – in contrast we have Immerseel with an absolut rigour of rhythm and without any dynamic response to the emphasis of the barline.

(Ravel bars 1-4)

What we have here is an invitation to dance by Ravel (as it seems to be normal in a dance like a „boléro“ – and in Immerseel’s case really an acoustical digitalised score!

For me the one big misunderstandig of Immerseel’s recording consists in confusing a dance rythm with mere mathematical strictness – and consequently of his understanding of the 20s and 30s of the last century. He seems not to have listened accurately but just followed the well known quote that „Ravel did’nt want his music to be interpreted, that it was enough to play it“, as we know from Alfred Cortot. Immerseel unfortunately didn’t consider that the performing tradition of Ravel’s time didn’t know the tempo contancy as a general rule, and Ravel of course just wanted to create a certain distance to this tradition – but which amount of distance? Well, his recordings gives us a very precise answer!

b. tempo rubato

There is a second and, in my opinion, very important aspect in Ravel’s version of the Boléro: it’s the tempo rubato, on the one hand as a collective flexibility of tempo, on the other hand as indiviual freedom. Immerseel excludes both aspects completely, his understanding of Ravel’s statement of tempo stability is an absolute one and doesn’t allow the slightest freedom. Ravel by contrast provides us with a lot of fine nuances: first of all he makes a clear distinction between the two sections, A and B;

In the A-section Ravel (in his recording) makes the musicians play in a very classical way, so very little rubato, not too much swing. No wonder, when you see the list of the instruments: flute, clarinet, oboe d’amore, some combinations...

tracklist / structure of Ravel’s Boléro

 1 Introduction (snare drum)
 2 theme A (flute)
 3 theme A (clarinet)
 4 theme B (bassoon)
 5 theme B (eb-clarinet)
 6 theme A (oboe d’amore)
 7 theme A (trumpet/flute)
 8 theme B (tenor saxophone)
 9 theme B (sopranino/soprano saxophone)
10 theme A (horn, piccoli, celesta)
11 theme A (oboes, clarinets)
12 theme B (trombone)
13 theme B (flutes, oboes, clarinets, tenor sax)
14 theme A (1st violins, flutes, oboes, clarinets)
15 theme A (violins, flutes, oboes, clarinets, tenor sax)
16 theme B (violins, trumpet, flutes, oboes)
17 theme B (strings, woodwinds, 1st trombone, soprano sax)
18 theme A (1st violins, trumpets, woodwinds)
19 theme B (tutti)
20 coda (tutti)

All instruments with a jazzy background play theirs solos in the B-section, saxophones, trombone, also the french basson which sounds a little bit squawky like a saxophone. And these musicians provide us with a much more emotional approach to the text and with the most expressiv rubatos. Most of them play the notes in the small circles quite slapdash (bars 7, 9, 11, 13), just ignoring that there are semiquavers in the score. And very remarkable is a collective tempo rubato at the melodic culmination on the dominant minor ninth cord (bars 5/6).

Listening for example to the trombone solo, we note the connection of articulation and glissando which is naturally very close for this instrument; todays trombone players, as far as I know, would never play like this. Unfortunately, I must say!

3 Ravel figure 10

Most amazing that Ravel’s trombone player plays the same rubati and glissandi also in the tutti, where it’s marked „sostenuto“ for him (after „15“ in the Durand-score). We can note as well very clearly the warm and slightly vibrating sound of the upper strings.

Ravel figure 15

As a contrast we can hear in Immerseel’s version of the same section strings playing non vibrato, trombone completely integrated; instead we hear a dominant soprano saxophone which apparently seems to be the only musician in the whole orchestra who was allowed to vibrate...

(unfortunately, we cannot provide you with a sound example for copyright reasons; see Zig-Zag Territoires, ZZT 060901)

As a conclusion I have the impression of a big misunderstanding. Once more todays musicians just read articles and books and don’t listen carefully to the most obvious evidences we have: the recordings of strong authority by the composers. However, this recording is not completely failed, there are many beautiful and interesting combinations of sounds, the balance is most transparent and thanks to the small string setting – just 37 strings – we can hear every colour of the winds, brass, harp and celesta. But we cannot really state that it successfully goes back to the spirit of the 20s as it intended.

2. The habits of the early music scene in comparison with early recordings

Now we are in the middle of the main subject: modern musicians – also today’s orchestras – go into the field of the tradition that is manifested through early recordings. And beeing one of these musicians I would like to bring up some practical details for discussion. For several years I have been principal oboe of the chamber orchestra of Basel (, an orchestra without principal conductor but with a few principal guest conductors. As many chamber orchestras currently do, we play classical music in a compromise version: gut strings, natural brass, period timpani and percussion but modern winds. In the field of romantic music we also did some Mendelssohn and Berlioz with Ophicleide and special horns. Generally spoken, we are accustomed to the rules of the game, called „historically informed early music practice“; we used to work with Christopher Hogwood as well as with Ton Koopman, Paul Goodwin, Giovanni Antonini, advancing to the romantic also with Paul McCreesh and Thomas Zehetmair.

After a few years under these conductors I realized that none of them had a complete conception about the eighty years between Beethoven’s death and the first orchestra recordings; one insists on original orchestral seating, another demands the strings not to use the spiccato because its use in the orchestra seems not to be historical, but asks for extensive dynamics at the same – which could not at all be realistic on period instruments; a third one thinks that the strictness of tempo and a consistent non tenuto playing is most appropriate to this period.

I actually dont believe in easy AND convincing AND historically indisputable solutions, and so I am most satisfied with the work of our really different conductors! They all understand that the rules of the classical music must have changed deeply during this period: it’s the transition of non-conducted orchestras to the modern conductors, in this context also the changing of stile of orchestral playing evidenced by writers such as Richard Wagner. In his polemic pamphlet „about conducting“ („über das Dirigieren“) he rages against the too easy and inexpressiv playing of music by and after Beethoven. His critique is addressed to the successors of, for example Mendelssohn, (in his eyes) mediocre capellmeisters and gives us strong evidence of different views and schools of conducting during this period.

See some dates to show how close some of the conductors of my research project are to Wagner’s tradition as well as to the Beethoven’s time. We can only assume, which stilistic elements we can detect in the Nikisch-recording go back to the „authentic“ Beethoven-tradition. We know that Nikisch was a successor of Wagner as a conductor with the aim of re-creating an interpretation and to work much with tempo modifications. On the other hand we know that already Czerny gave the advise to take a slightly different tempo for main and second theme of a sonata. In any case I am sure that with some recordings (eg. Nikisch, Strauss and Ronald) we have a wide open window to the 19th century.

This gap seems to be exactly the same conflict as we can detect between todays early music scene and the habits of the orchestras 90 years ago. We really can detect some clashing elements, as we stated before in my Ravel-introduction, a fact which is of course not surprising! Let’s have a look at this gap and let’s consider if it’s just the 80 years of history that make the difference or if it’s indicating the degree of error of the early music scene.

This is a very short overview which of course doesn’t cover all regional and individual differences, it just shows some clear trends in early recordings as well as in the modern early music scene. The more intensiv the colour becomes, the bigger is the collision between the two stiles.

There are no big problems with the different use of vibrato, I suppose, we have none or just a few on both sides. A bit more with articulation and fidelity to the original text; a certain importance we note regarding the size of orchestra, because there changes a lot when you enlarge an orchestra, but we have to consider that there were big orchestras with double winds already during Beethoven’s time.

On the sostenuto-playing: there is no big fight about the fact, that Beethoven defines an exception of a rule of his time when he writes „tenuto“; so it’s very likely that the early music scene is historically right with its more or less non tenuto playing. Anyway, the changing of the classical to the modern bow which was the beginning of the tenuto playing fells into Beethoven’s lifetime, so there must have been a fundamental change of style already in the first half of the 19th century. So I would suppose that Roger Norrington is certainly historically wrong, when he plays Mahler’s music in a almost baroque manner without sostenuto phrasing.

For Beethoven’s music the real big fishes are the three last aspects!

a. Portamenti
With the playing of portamenti today’s musicians really have quite a big deal of a problem, because it’s a fact that this tradition seems to be over – at least as a appropriate stilistic device for expression in the classical orchestra. I can remember a situation with Christopher Hogwood, when we were recording Haydn’s Sinfonia Concertante (released with the „kammerorchesterbasel“ on arte nova) and our concert master made a wonderful portamento in the style of a romantic virtuoso. His reaction was very dry: „we are not this kind of players, are we?“ And in his face it was written very clearly how he thougt about „this kind of players“, and so in the end we took another version... There are indeed some convincing portamento soloists as Daniel Hope or Pieter Wispelwey or chamber music groups as the Eroica Quartet. In the field of orchestral music you all probably know the recordings of the Smithsonian Chamber Orchestra with Barber and Mahler’s Adagietto; honestly and with all respect, I’m getting the feeling of imitation more than of the expression of a real emotion, listening to this group. The original (in the case of Mahler the Mengelberg-recording) is much more moving, even a bit messy and uncoordinated – which makes probably the difference! Definitely we cannot separate the idea of portamento from its emotional content and just execute it as a rehearsed habit. Perhaps we did not yet reach to the point of more adventure or it’s simply not the natural performing-stile of today’s  musicians.


b. tempo/stability of tempo

Giovanni Antonini is the one of our conductors, who are aware of the tradition of the orchestral playing of 90 years ago, the other one is Thomas Zehetmair. The increasing number of conductors of this kind by the way shows sort of a trend for early recordings becoming evidence of a reliable relevance. Together with Zehetmair for example we performed Brahms’ first symphony some weeks ago and precicely followed some descriptions of an authentic performance of the 19th century by Steinbach and on this base he worked out a most interesting concept of flexibility in tempo and rubato-playing of individual instruments (Sir Charles Mackerras did something similar already in the early 90s); in the end an interpretation turned out that was not so different from Mengelberg’s Amsterdam-version (1940), a recording which Zehetmair by the way didn’t know!

Johannes Brahms: 1st symphony, excerpt of mouv.4 – Live recording 21-8-08
Kammerorchester Basel – Thomas Zehetmair

Sometimes I really wish that modern conductors would go much further than they go in order to provide us with exiting performances and not just with nice and clean ones. Try to imagine this would happen today e.g. at the Proms in a concert of the London Philharmonic:

Beethoven: 5th symphony, transition 3rd to 4th mouvement
Royal Albert Hall Orchestra – Sir Landon Ronald (1926)

A very unconventional solution of this transition, also for his time an extraordinary powerful one...


c. rhythm / tempo rubato (vertical precision)

As we all know the vertical precision is a sign of quality of an orchestra. Our ears and the ears of the public are CD-accustomed and have a expectance of perfection also in concerts. The amount of technical achievement in today’s orchestras has risen to a level that never existed before. Can we be sure that in the following example the musicians really played what they intended to? Is there also a certain degree of haziness and does it matter if there is?

It is probably one of the most famous rubati in the history of recorded music, an episode from the 1st mouvement of Beethoven’s 3rd piano concerto.

Beethoven: 3rd piano concerto, beginning of the transition of mouv.1
Marguerite Long – Paris Conservatoire Orchestra – Felix Weingartner (live 1939)

If that would happen exactly like this in a today’s concert, you might be sure to read in next day’s newspapers, that it was an expressive interpretation but unfortunately not rehearsed enough. This could be the main reason for the impossibility of turning back to the habits of the early recording era: we captured the sound from the reality, from the concert, from the human beeing and so it happened that the music became just sounding digital information.

Perhaps we can also detect another crisis behind this process, namely a crisis of repertoire. Is it really indispensable to play the same symphonies over hundreds of years? In the field of the contemporary music it is eventually completely different: it happened many times during my life as performer that I played a new piece just one time, and that was it; this is the contrary extreme, of course, but I feel sometimes much better as a musician when playing music that is a new experience for everyone attending in the concert hall and to be not just a reproducing machine creating memories.

A last example for extensive use of rubato to end with this subject: In the trio of the Scherzo we have a fugato which gives Landon Ronald the inspiration for a most exciting rubato, marked here with arrows. You will hear a remarkable conflict between crotches and quavers. It’s quite dangerous to play like this, but it’s fun to listen to, as I think...

Beethoven: 5th symphony, trio of the 3rd mouvement
Royal Albert Hall Orchestra – Sir Landon Ronald (1926)

d. Fidelity to the original

To end with, I would like to touch the subject „fidelity to the original“. When I was in Vienna some months ago I had a day off before a concert and decided to look for Mahler’s version of Beethoven’s Fifth; I had known for a long time, that such scores existed, but I was surprised to find very easily also the complete performance material. To estimate Mahler’s achievement one has to be aware of the fact, that he was just responding at the changed conditions of orchestras and concert halls during the last 50 years. So he doubled all woodwind and wrote two additional horn parts – since also the strings had been duplicated to about seventy in the big orchestras that he was conducting (Vienna, Amsterdam, New York), just in order to balance out Beethoven’s structure. First of all he defines clearly where the playing of the winds is soloistic and where doubled (there are original markings for this by Beethoven himself for the 4th symphony!).

Secondly he specifies all dynamics much more detailed than Beethoven did, throughout the entire symphony – as you can also see on this two pages, probably also considering the few rehearsals he got and makes in a third step some changes of instrumentation where he has the feeling that the balance is not fine enough.

In the following example from the second mouvement the oboes play together with the flutes to have a stronger line against the violins; at the same time the second violins play at the same octava with the first, sustained by melodic trumpets. As a matter of facts there really are substancial variations to Beethoven’ original.

Since Mahler not considered with recordings – his version was elaborated in the 90s – he was just interested in a big and stunning effect of the performance in big halls, no wonder that many conducters after him applied some of the ingredients of his arrangement even until the second half of the twentiest century. On youtube one can see Karajan using even 8 horns for this symphony!

Why I would narrate about this so detailed? Integrated in my research project in Berne we plan to realize this version in concert, probably in spring 2010. We think that Mahler’s score is a outstanding evidence of the view of a very modern conductor in his time – he was considered even the best! – who made no much difference between his own works and the works of other composers, and let us participate in his way of interpretation. Unfortunately he didn’t write many agogic details in his score and so we have to go back to the numerous descriptions of his stile of conducting to find our way to an interpretation in the spirit of the fin de siècle. I found one most interesting indication in the score of the Vienna Philharmonic Orchestra, where he conducted the piece twice in 1899.

The archivist listed all performances between 1844 (Otto Nicolai) and 1929 (Franz Schalk), until the material must have been changed. Apparently the timing was interesting enough that he decided to write it down, 36 minutes – well understood without the repetition in the fourth mouvement.


Watching the list of several timings from Nikisch to Gardiner we can easily note, that the planned performance will be far away from Beethoven’s original metronome markings of 1817! There are only two conductors which are even slower than Mahler (presumably) was: Hans Knappertsbusch and Peter Tiboris. The latter by the way made his recording incorporating Gustav Mahler´s „Retuschen“; he reconstructed the orchestra size of the 1900-New York Philharmonic; his version however turned out quite ambivalent: heavy and very strict tempos (we cannot feel the freedom of tempo which apparently must have been Mahler’s style) and a modern orchestra playing (permanent vibrato, no portamenti, strong and heavy articulation). I have no doubt that we will discover much more of the spirit of Mahler’s interpretation considering all relevant aspects of style of his time, as we know them from early recordings.


© 2008 by Matthias Arter




Convergence Youtube FacebookTrashImpressum