Ben's Rite of Spring Blog Entry
Posted: 2010-04-16 11:41:00
THE TWENTY-YEAR STORY OF THE BOSTON PHILHARMONIC AND THE RITE OF SPRING
Benjamin Zander
The story begins in 1990 with our first performance of the Rite of Spring in Jordan Hall, which was captured on an extraordinarily vivid recording by Ken Dean. The following year this was issued by IMP Masters in England (http://www.mediafire.com/?sharekey=02c449548637afd46e7203eb873681298e14e43cc7bbb456759e682a8cd2154a). With the recording of the performance of the Rite went a recording of a piano roll “realized” by Rex Lawson and thereby hangs a tale.
RIGHTING THE RITE
It is little known that in 1920 Igor Stravinsky began supervising programming of piano rolls for the major works he had so far composed, "in order", as he said "to create a lasting document which should be of service to those executants who would rather know and follow my intentions than stray to irresponsible interpretations of my musical text".
The most astonishing aspect of the rolls of the Rite of Spring is that the final Dance Sacrale, in which the young virgin dances herself to death, goes far faster in relation to the rest of the work than we ever hear it played now, even on Stravinsky's own recordings.
What happened? Can Stravinsky have indeed intended this to be the tempo of the section when all his scores bore a metronome mark nearly 30 points slower than what we hear on the piano roll?
The eminent American musicologist, William Malloch, first drew my attention
to this remarkable information. One of the most insightful and creative thinkers
about music today, Mr Malloch has explored in great depth the way in which the
existing technology of the day-i.e. barrel organs, music boxes, piano rolls and
metronomes - can shed light on the tempi at which pieces of music were
originally played. He argues that the reason Stravinsky reduced the tempo for
this section was that the music was so difficult for the players of the day and
indeed for Stravinsky himself to conduct, that the composer simply wrote in the
score and used in his own performances a tempo that he and his musicians could
manage. All conductors since then have followed suit.
It is interesting to note that Pierre Monteux, the man who conducted the work
at its stormy premiere in 1913 and therefore the musician closest to the
original conception, is heard struggling to drive the Paris musicians into
playing the section at very nearly the tempo of the piano roll on his 1929
recording, though the results verge on chaos. Stravinsky fares no better in his
own recording of the same year at the slower tempo, underscoring the fact that
this music presented an extraordinary challenge for even the best European
musicians of the day at any speed!
Malloch ingeniously offers the theory that other composers who happened to be
in Paris around this time were influenced by the heady effect of the Dance
Sacrale and quoted it in their own works.
In Bartok's Sonata for Two Pianos and Percussion, Aaron Copland's the Young
Pioneers and, most strikingly, in the Finale of Prokoviev's Seventh Piano
Sonata, there are passages that quote almost verbatim the Dance Sacrale, but at
the tempo of the piano roll, not at the familiar slower tempo. Since all three
composers, by coincidence, made piano rolls of their own works at the Pleyel
studio it is inevitable that they would have heard Stravinsky's roll of the
Dance Sacrale at the fast tempo and it apparently made an indelible impression
on them.
Could the dancers have danced the section at this breakneck speed? Perhaps
this wasn't Stravinsky's primary concern. At one of the rehearsals for the Rite,
Marie Rambert, the great Russian dancer, describes how "Hearing the way the
music was being played, Stravinsky blazed up, pushed aside the fat German
pianist ... and proceeded to play twice as fast as we had been doing it and
twice as fast as we could possibly dance. He stamped his feet on the floor and
banged his fist on the piano and sang and shouted
At the galvanizing speed of the piano roll, the conclusion of the work's
second part matches and even surpasses the cumulative excitement of the first
part, instead of being something of an anti- climax and certainly, we can
readily understand how a sacrificial figure could have danced herself to death
at such a tempo, whereas at the slower tempo she seems to have a chance to
survive!
No one will ever know what Stravinsky would have wanted had he heard the
Dance played by an orchestra at the tempo of the piano roll. It is hard to
imagine that he would not have joyously embraced it as "the truth", so
staggering is its effect.
Perhaps there is poetic justice that a conception that was musical training,
to study and perform the great works of the abandoned because the most
experienced musicians of the day orchestral repertoire and to work under the
inspiring leadership could not realize it in 1929, is finally restored to its
original power of founder Benjamin Zander. by a semi-professional orchestra 70
years later!
Here is some crucial information about piano rolls in general and the Rite in particular provided by Rex Lawson, which appeared in the booklet.
Paradoxically, part of the attraction for Stravinsky of these
metronomic rolls was his belief that they helped him to limit
excessive freedom on the part of concert artists, and to that intent
he was keen to specify roll speeds and general dynamic changes.
By the early 1920 his acquaintance with the pianola (Etude pour
Pianola - 1971) had grown into a deep involvement with the
instrument and in particular with the Pleyela, manufactured by
the French firm of Pleyel, Lyon et Cie. In his memoirs Stravinsky
recalls the fascination he had for the arrangement of his works
directly for music roll, where the limitations of ten fingers and two
hands are simply irrelevant.
Working in an apartment which he rented in the Pleyel building
in Paris, he was able to maintain direct control over the
translation of his works on to roll, a task which he delegated in
detail to Jacques Larmanjat, the head of Pleyel's musical
department. In this way, over a period of about six years,
Stravinsky arranged most of his major ballets especially for the
Pleyela; Petrouchka, Firebird, Les Noces, Pulcinella, Song of the
Nightingale and of course, nine rolls of the Rite of Spring.
Although he did not physically record the rolls, which have in any
case far too many notes for a human being to manage, he was able
to specify the number of perforations per beat in each section of
the music, giving us an exact reflection of his ideas of tempi at the
time.
None of Stravinsky's Pleyela rolls has even been recorded on
disc before, and on this occasion a 1912 Aeolian push-up Pianola
was used, a device that fits in front of a normal piano and plays it
by means of a set of felt-covered wooden fingers. The piano was
a Bosendorfer Imperial concert grand in the Great Hall of
Dulwich College, London.
-Rex Lawson
The reviews for the disc, and especially of the Danse Sacrale were uniformly enthusiastic and the sales, for a CD, considering it was made by a semi professional orchestra and a barely-known conductor, were, as the saying goes, “off the charts”. The high-water mark was rached when John Rockwell named the recording in the New York Times’s end of year round-up, as one of the ten most important musical events of 1992.
First a review from Classic CD which compared nine different recordings of the Rite and accorded the Boston Philharmonic recording more stars than any other (including Stravinsky’s own!)
-Classic CD
And here are some of the critical responses in the Boston and international press:
"Mr. Zander's orchestra plays with exhilarating authority. The Finale is positively harrowing: cross rhythms explode against one another with brilliant virtuosity. That the performance was taped live makes it doubly impressive...the sound is superb, the interpretation likewise."
John Rockwell - The New York Times
"...One of the greatest recordings of the year... the recorded sound will take your breath away... an audiophile's dream... a performance which would do any orchestra proud, a thoroughly convincing interpretation and one of our greatest Sacres..."
Fanfare - Leslie Gerber
"...One of the most extraordinary discs for some time...has to be heard to be believed...the performance has a blazing intensity that few of its slickly produced studio rivals can match...You could not say you truly know the work without it."
Hi-Fi News
"Stunning, positively primeval...with a breathtaking finale...brilliantly transferred to disc."
The Boston Phoenix - Lloyd Schwartz
The Rite of Spring remained dormant in my life for ten years, till I programmed it again for a Boston Philharmonic concert in 2000.
Then one of the members of the orchestra drew my attention to an article by a scholar called Professor Robert Fink. In a footnote at the end, our recording of the Rite was mentioned:
That is why the only “historically informed” recording of the Rite to date remains an interesting
but ultimately failed curiosity. Conductor Benjamin Zander thought to achieve “authenticity” by reproducing the exact tempos from the Pleyela piano rolls in performance with the Boston Philharmonic (IMP MCD 25 1989)….
he simply chose the fastest tempo of the Danse Sacrale he could find. Even more symptomatically, he ignores the actual tempo fluctuations on the roll. Once he has his new fast tempo, he holds to it grimly and geometrically to the end. If we are going to hallow these piano rolls as evidence of the composer’s intentions, ought we not listen to them all the way through?
Of course I was irritated at first to be hauled over the coals like this – “I had so listened to the whole piano roll through”, and hadn’t all the critics heaped encomiums on our revolutionary interpretation praising us to the skies for restoring Stravinsky’s true intention? But remembering Rule # 6, I read on. What Professor Fink presents in his article is nothing short of a musicological bomb shell. What he argues is that Stravinsky began as a romantic (a vitalist, in Fink’s terminology) and that he moved increasingly towards what Fink calls a geometric “rigoroso” approach to the interpretation of his music, and, according to Fink (and Richard Taruskin), influencing along the way, all musical interpretation of the authentic period instrument movement. In simple terms this meant that whereas in the earliest performances (and recordings) the Rite was approached as a work of “romantic” music with high emotionality, tempo fluctuations to underline the drama, rubato etc, it became increasingly approached as a work of clean, sharp lines and sterile Affekt.
Here is the Abstract of Professor Fink’s article:
It is only recently that we have begun to consider modernist performing style-especially its brisk, unyielding tempos and abhorrence of "expressive" rubato- as a historical phenomenon. Much of the credit (or blame) for this style has been ascribed to the composer of The Rite of Spring; Richard Taruskin argues that "all truly modern musical performance treats the music performed as if it were composed-or at least performed-by Stravinsky." But the performing history of the Rite shows that the composer struggled mightily to get his own music played "as if composed by Stravinsky. "Early interpretations of the Rite were slower and more elastic-more "romantic"-than the composer wanted. Focusing on the "Danse sacrale," this paper examines the battles over tempo and rubato evidenced by historic recordings, piano rolls, and published documents. It also considers the unpublished compositional and performing materials for the Rite: Stravinsky's autograph short and full scores, and his annotated personal copies of the 1913 piano reduction and the 1922 and 1948 fill scores. The record indicates (1) that tempo and pacing of many sections of the Rite were radically rethought between sketch and 1922 printed score; (2) that someone (Pierre Monteux?) indicated rubatos and changed many of Stravinsky's metronome marks on the autograph; (3) that early performances of the "Danse sacrale " featured unwritten tempo modifications for dramatic effect; and (4) that Stravinsky had to work for decades to fix in his score the rigoroso that has become the characteristic performing tempo of our time.
Here is a link to Dr. Fink’s entire essay: http://www.benjaminzander.com/uploads/files/Fink.pdf
Much to his surprise, I called Professor Fink (wasn’t I, after all, a member of the enemy camp?) and asked him to guide me in preparing a vitalist performance of the Rite for our 2000 concert. We had several conversations and I absorbed many of the ideas he had put forward in the article. The result was a much more convincing performance (http://www.mediafire.com/?sharekey=02c449548637afd46e7203eb87368129b0e5ff8f218ca5f3a9a26c4ed87536eb).
Now, ten more years later, and before preparing another performance, I read his article once more and approached him again.
The correspondence we have had over the last few weeks is printed below.
***************************************************
Benjamin Zander wrote:
Dear Robert,
I have just re-read your brilliant and illuminating article on the Rite, because tomorrow I start rehearsals for my third performance of the piece with the BPO on April 22/23/24/25
(1990; 2000; 2010 seems to be revealing a pattern!).
I will be more aware than ever before of the vitalist opportunities. You can be sure that it will be the guiding principle of my work with the orchestra over the coming weeks.
Perhaps the outcome will be a recording that you can hold up as an example of the "true" spirit of the Rite that you have so eloquently revived.
I wish only that you could be here to help guide us in our endeavor.
I remain a true believer and admirer
Ben
***************************************************
Dr Fink responded:
Ben
Thanks so much...if you do get a recording you like, I hope you will share it with me. :)
************************************************
A few days later:
Dear Robert
I listened to my 2000 recording yesterday - I find I get so busy that I rarely listen to past performances - and found it quite satisfying. I believe
I sent you a copy. If so, would you be willing to listen again and suggest ways in which we might be able to more completely embody the vitalist
approach to the piece? I would find that extremely valuable.
Also do you have the article in a form that I could post it for my Graduate Interpretation Class. Perhaps you could email it to me?
**********************************************************************
Dr Fink responded:
Ben
I have to agree, that 2000 recording holds up quite well! On re-listening, I found that the Wagnerian "pure allegros" really work, especially in the Danse sacrale and the Glorification of the Chosen One. One thing I quite like is the way you phrase the more "lyrical" melodic moments in the score, especially when they are scored for strings - like R61-62, for example, or the opening of the Mysterious Circles. (I also liked the ever-so-slight rubato from the alto flute at R93 - I wish the clarinets had followed suit and loosened up a little...)
I may have more to say on reflection, but I would concentrate on two tempo issues in the score that would have a huge vitalist effect, and which I still am looking to hear:
I. Danse sacrale
a. I think the tempo you got for the opening A section of the Danse sacrale is fantastic, and I heard you take a slower tempo for the "B" material at R149. But I wish there had been more of the accelerando at 2 before R167 - this whirling music in the violins is associated with the Chosen One spinning around faster and faster, and then freezing in place at the fermata, so the feeling of a machine spinning out of control would be wonderfully appropriate. AND it would naturally get you back into the faster tempo for the return of the A material at R167.
b. I can also hear the distinct slowing down at R174 - but I really think it could be more! If it feels wrong to jam on the brakes too hard there, then I would plead with you to try to make R181 noticeably slower than R174, and even heavier (it seems to be scored that way). Remember that the idea of the dance is that the Chosen One is continually tiring, especially thanks to the huge clumsy movements she is forced to make during the C section - so it is dramatically plausible that R181 is more slow (= more exhausted) than R174. Also I wish you could get some of the beautiful tone that you got from the violins at other places during R174-180 and R 181-186; in general, I think a vitalist (Monteux-style) reading of the piece would still hold onto that old - but not quite dead - conductor's idea that the first violin line is the default "Hauptstimme" of the piece. I would love to hear the violins singing out from R181-186, even though the brass is blaring and trumpeting (maybe they could be persuaded to play a little crisper?); that violin ascent from D to G is the linchpin of the whole piece's ascent to the climactic final high A. (In the 1943 revision, Stravinsky added horns to the violin line, and thinned out everything around it...)
c. You really got the sudden jump to "galop" tempo, and very tight it was, too (better than Igor did with that orchestration)! But before I die, I would like to hear a modern orchestra with the discipline of yours do the passage as Stravinsky first wrote it in 1913 (and as it is in the 1921 score, and in the Monteux and Stokowski recordings), without the trombones and tuba oom-pahing away like a circus band. There would be a real panther-like menace if you could get a sudden sfff-pp at R186, which is impossible to do as Igor re-scored the passage in 1929. I could try to find a 1921 score for you and see how many bars would need to be rescored - I bet if you just took away the trombones and tuba between R186 and R189 you'd basically have it...
II. Augurs printanieres - Danse des adolescentes
a. This entire section (R13-37) is a little stiff in your 2000 recording. In the oldest recorded performances, the whole section accelerates, so that it is traveling at quite a clip by the time it ends, noticeably faster than the opening chords.
b. I'm sure this would stress out an orchestra, but I'd love to hear it done the way early recordings do it. I imagine that for the first generation of conductors, those endless eighths in 2/2 meter looked like Beethoven's Fifth. (Bear with me...) And like the Fifth, it seems that conductors gave themselves permission to slow down whenever the "fate knocking on the door" accents came in. (OK, in this case Fate is a little tipsy, and knocks irregularly...) So I would start R13 slower than the main "allegro" tempo of the section, then let things naturally accelerate; then deliberately hold back at R18 when the opening accents come back; then let things accelerate and reach cruising speed by R28.
c. Now comes the key part, which will really knock an audience back on its heels. The way one can let things speed up is through agogic accents when the horn offbeats come in. In early performances, the orchestra really punches those accents - at R13, the eighth note pulse starts under the eventual main tempo, but each time the horns attack, that beat is just slightly early, like a quick punch that hits the body just a little sooner than you expect. And every time a beat comes in early, it speeds up the tempo just a hair. (Note that in the ballet, every one of these accents is a literal quick punch into the air by the men.) Once you get to about R28, things are pretty geometric; but at the opening, the eighth notes don't have to be a rigidly regular grid, any more than the same eighth notes in Nikisch's famous 1912 recording of Beethoven's Fifth are all equally spaced. (Metaphorically, one could imagine a perfect Cartesian grid of eighths, but accept that those heavy accents impose a distortion on the grid, like gravity does in Einstein's way of understanding space-time. In a way, every time the horns add their weight to the beat, they create a little gravity well, into which the piece can accelerate.)
I know this is much harder than the metronomic eighth notes you hear in contemporary recordings; but the Rite is not a Haydn symphony!
Anyway, it's something to think about...
rwf
*************************************************************
Robert,
I am extremely excited by what you write. Thank you so much for all your trouble.
I would love to see the 1921 score to see what he actually had there. If you have access to it, it would be very much appreciated.
The trouble with just leaving out the trombones and tuba is that the top voice isn't represented (i.e. the E and G and the A natural on the second 1/16th note). Did those notes not exist in the
1921 version? I cannot pick it out from the murk of the 1929 Monteux recording.
Everything else you suggest will be put into practice at the next rehearsal ( am especially excited about the idea of starting Augures Printemps slower and speeding up and coming early with the
horn accented notes. It sounds great in my head!
Thanks to you, we are inching towards a vitalist modern recording! ((A little leap every 10 years, isn't bad!).
What is the 1943 score you refer to? Does he actually change the horn line to support the violins? Who takes the canonic writing that now is in the horns?
We are playing from the 1947 score which was revised in 1967)
Best wishes
Ben
***************************************************************
Ben:
I found my copy of the 1921 first edition of the Rite, and have scanned the key pages (R187-189) for your delectation. It turns out that it is not quite as simple as taking the trombones out - the offbeat chords are played by bassoons + low horns, in a different voicing that gets all the notes of the chord, and I think the string parts are a little different, too. (In this version, the bass drum part becomes an "obbligato," since it will be clearly heard.) We're basically talking about eleven bars, before the 1921 and 1929 editions converge.
If you wanted to make this change, I have uploaded scans of the relevant pages to my filesharing page. I scanned the score pages, and then did detail scans of the string parts and the bassoon/horn lines, so you can read them.
[These can be found at http://www.benjaminzander.com/uploads/files/1921_Rite.pdf
The only other fun difference is the noisier percussion at the very very end. I have scanned the final bars; note the part for "guero" (rasp)! If you could make that heard, the final crash-bang will sound a little more exotic, to say the least.
Hope this comes in time to be of use...
rwf
********************************************************************
Benjamin Zander wrote:
I would love to play the eleven hammer blows slowly, with overwhelming power. Monteux doesn't. Can you think of any justification other than that it would be the ultimate vitalist gesture?
I am excited to try out all your ideas.
Best
Ben
I like it. All there is on the scores is (uselessly enough) dueling divisions of the 11 into 3's and 4's - (Stravinsky and Monteux do it differently.) No tempo modifications there AFAIK. (I am away from my materials, but I can check what is happening on stage at that point on Sunday.)
I found that part of your performance very exciting.
**********************************************************************
Dear Robert,
We tried the 1921 scoring of the Danse Sacrale last night (we had to write out all the parts). I am afraid we couldn't make it work in the short amount of time we had. I am disappointed that I can't give you a nice version to listen to, but you will have to wait, sorry.
I see why he changed it to the low brass. It's much easier to lock into the new tempo.
We aren't back up to the MM152 yet, but intend to be by the concert.
It's all very exciting and VITALISM certainly lives at the BPO.
People are struck how lyrical and gorgeous so much of it sounds. Your name is bandied about quite a bit around here these days.
More soon
Best wishes
Ben
For some more interesting material on the BPO’s journey with the Rite of Spring, read Richard Taruskin’s chapter on our recording, along with my response:
http://www.benjaminzander.com/uploads/files/Taruskin_Stravinsky_Lite.pdf
And my response:
http://www.benjaminzander.com/uploads/files/BZ_response_Taruskin.pdf
Summing up:
I think that perhaps the most interesting single clue for the argument for a vitalist approach to the Rite, is the 1929 recording by Monteux. He was, after all, the conductor at the original performance and had led the early rehearsals. I had heard it in 1989 and had taken note that the tempo he takes for the “Danse Sacrale” is just about the same as the piano roll, suggesting (along with the evidence of the piano roll itself) that, in all probability, that was I.S.’s original conception. What I hadn’t noticed, however, through the fog of the terrible recording and the quite shoddy playing, was the myriad of minute tempo modifications throughout the performance. If you listen to it again you hear a romantic vitalist composition not a geometric, sterile clipped rendition.
It’s is rather an amazing state of affairs. Our performances on April 22, 23, 24 and 25 will go full out for the vitalist approach. No holding back this time!
