Our Summer BAL 60: Stravinsky Symphonies of wind instruments

[Pulcinella]

Stravinsky's Symphonies of wind instruments, dedicated to the memory of Claude Achille Debussy, who died in 1918, was composed in the summer of 1920, and first performed in London on 10 June 1921, conducted by Koussevitsky. Early sketches date from 1919 and are scored for harmonium. A piano version of the last section (some 50 bars or so) was published as Fragment des Symphonies pour instruments à vent à la mémoire de Claude Achille Debussy in a music supplement of La Revue Musicale, December 1920, entitled Tombeau de Claude Debussy, where it appeared as number 7 in a set of 10 compositions by different composers. The full score was not published, but a piano reduction made by Arthur Lourie was published in 1926. Stravinsky later revised the work, and the revised score was published in 1947, and it is in this form that the work became widely known. A performing edition of the original 1920 version was published in 2001, and this version has been performed (and recorded) with increasing frequency.

List of available recording (courtesy Presto website).
The version is given where identified and timings are given [in square brackets]; if there are two timings, then the recording appears in more than one incarnation, with different track timings.

Ansermet: OSR [9.14]
Ashkenazy: Berlin RSO (1920 version) [8.59/9.07]
Bergby: Royal Norwegian Navy Band
Boulez: BPO [9.16]
Boulez: NYPO (1920 version) [9.38]
Boulez: Orch Domaine Musical [8.52]
Boulez: Orch Nat de France [9.33]
Colburn: President's Own United States Marine Band [8.23]
Craft: Columbia Sym winds and brass (1947 version) [8.35/8.41]
Craft: Twentieth Century Classics Ensemble [7.59]
de Waart: Netherlands Wind Ensemble [9.08/9.13]
Dutoit: Montreal SO (1920 version) [9.04]
Fennell: Eastman Wind Ensemble [8.41]
Fischer: Netherlands Wind Ensemble [9.10]
Foley: President's Own United States Marine Band [8.24]
Graham: US Air Force Heritage of America Band [9.59]
Jurowski: LPO [9.04]
Nagano: LSO [10.16]
Rattle: BPO [9.46/9.54]
Rattle: Nash Ensemble [8.58/9.04]
Salonen: London Sinfonietta (1947 version) [8.41/8.47]
Shao: NZSO [8.51]
Stravinsky: WDR Koln (recorded 1947) [9.11]
Stravinsky: Cologne RSO (recorded 1951) (1947 version) [9.29]
Stravinsky: SW German RSO (recorded 1954) [10.18]
Van Zweden: Netherlands Radio Chamber Phil [9.08]
Welser-Möst: LPO [9.10]
Winnes: Swedish Wind Ensemble [9.48]
Wyman: US Coast Guard Band [8.49]

Piano arrangement by Lourié
Jones [7.58]
Koehlen [8.56]

[Pulcinella]

Some additional resources/background information.

(1) Video with score (posted by teamsaint on the YouTube thread):

https://youtu.be/72V_OTSiAkM

London Sinfonietta
Conductor: Esa-Pekka Salonen

(2) Tempo investigations
A study of adherence to tempo indications in some recorded versions:

http://davidvickerman.net/stravinskys-symphonies-recording-tempo-discoveries/

(3) An article entitled Unity in Stravinsky's Symphonies of Winds by Brian Diller:

http://www.briandiller.com/uploads/5/7/7/2/57728231/stravinsky_symphonies_of_winds_final.pdf

[Pulcinella]

More background material (kindly supplied by fernyhoughgeliebte).

The chief sources were Robert Craft's chapter on the work in his Stravinsky: Glimpses of a Life, and the relevant pages in Stephen Walsh's Stravinsky: A Creative Spring. The former's from 1992, the latter from 1999 - so there may well be Forumistas with more up-to-date information.

The work had a rather complicated (!) pubication history. Within the first three years of its existence, it was performed seven times: Koussevitsky (10th June '21); Goossens (12th December, '21); Ansermet (26th Dec, '21 - Ansermet had a bit of a "thing" about Boxing Day performances of the work: the next time he performed it was 26th Dec '22); Stokowski (Nov '23, and Feb '24); Prevost (Jan '24) - after which the composer forbade any performances except those conducted by Ansermet, who programmed the work twice in 1928. Although a possible publication of the score was mooted several times, the composer didn't seem particularly moved to press for this - unusual for a composer whose custom it was to badger publishers to get his scores in the shops as soon as possible.

Stravinsky's preference for Ansermet wasn't solely due to the quality of that conductor's Musicianship (indeed, he never heard Stokowski's performances, and it's not clear if he was present at Goossen's - he was at Prevost's, which he found far too mild-mannered) but more to do with the number of errors present in the materials that those other conductors used. (It's difficult to follow - but I THINK that there were three sets of parts and conductor's scores around, copied by different people at different times - and the either the composer no longer had his own manuscript of the work, or he didn't consult it later.) Ansermet's copy also had errors - but he was in closer contact with the composer, and could ask questions about possible errors.

In 1933, in readiness for a possible publication of the work (which never happened), Ansermet corrected the proofs of the score from his own copy. (The composer was too busy working on "Persephone" and passed the work onto Ansermet.) Ansermet sent regular letters to Stravinsky querying matters of detail (usually of the "does the bassoon in bar x keep an F sharp in the last quaver, or does it change to a natural to double the clarinet"- type questions) - SOME of which the composer replied to; others seem to suggest that the composer of "Persephone" wasn't entirely sure what the composer of the "Symphonies" had wanted! The complete 1933 Ansermet edition is an impressive feat of loyalty and service to a composer, but it still contains several errors/discrepancies when compared with the original score. BUT - and this is where the fun really starts! - it was this version, in Ansermet's handwriting, that Stravinsky used in 1947, when he came to revise the work - AND some of these "errors" were simply copied by the composer in this revision. After completing his 1947 work, Stravinsky instructed his publisher to withdraw the earlier materials from circulation (which didn't stop performers who preferred the earlier version from getting hold of them).

Robert Craft (whose request in 1945 for permission to perform the work prompted the composer's urgent wish to revise the work - and inaugurated their quarter-century relationship) says "to compare the original and the versions is a bewildering exercise ... The instrumentation is entirely different, and so is much of metrical structure, the phrasing and articulation, the rhythmic figuration, and even the harmony. Th most striking of the 1947 revisions in rhythm is the substitution of quavers for triplets in the first Tutti ... But the 1947 version weakens even the bell motive, with comparatively colourless instrumentation and the reduction in the number of 'peals'.

===

"Above all, the original is richer and more varied in instrumental colours. Consider the flute and clarinet combinations, without oboes, in the four bars before fig 35, at 38, and at 41 - in contrast to the sharply accented and detached chords of the oboes and bassoons in the 1947 score - ans well as the mellow trumpet of the last phrase before the final tutti, and the quiet open brass of the ending vs the nasal muted brass of 1947. The softness of these sonorities, enhanced by legato phrasing, provides a change of feeling and a sense of progression absent in the 1947 score."

Craft's comments are particularly in response to the preference for the 1947 score over the original expressed in the video linked to "above" (ie the three-part documentary in post #4). But he concedes that the two versions are "'fundamentally different', the result of entirely different aesthetic approaches". To get the work as the composer imagined it in the fire of its first thoughts, a recording of a version taken from the composer's original manuscript score - with Ansermet and others' errors edited out - is essential. But the 1947 revision is a remarkable insight into how the composer's imagination had been transformed by the experience of Neo-Classicism - and, just as the work was originally a "farewell" to the "Russian" period, just before he "moved into Neo-Classicism", the 1947 can be heard as the beginnings of the composer's farewell to Neo-Classicism as he moves towards a Serial imagination.

[Pulcinella]

Another set of links from ferneyhoughgeliebte:

Incidentally, I [ferneyhoughgeliebte] posted youTube links on the Forum some weeks ago to a rather splendid 45-minute documentary, "The Last Chorale", which is full of fascinating information. If you missed it -

https://www.youtube.com/watch?v=eHGbBw12Dc4

https://www.youtube.com/watch?v=IMTG2zYcYFQ

https://www.youtube.com/watch?v=x_2vSduqLyQ

[verismissimo]

Good choice, Pulci, for our first of this summer. And fine list!

At least, in this part of the empire, we don't have to complain about twofers, my fav being ignored, presenter bias etc etc.

[Pulcinella]

My own preliminary thoughts and comments.

Here is a list of the 11 versions I found that I had on my shelves.

Ansermet: OSR
Ashkenazy: Berlin RSO (1920 version)
Boulez: BPO
Boulez: NYPO (1920 version)
Craft: Columbia Sym winds and brass (1947 version)
de Waart: Netherlands Wind Ensemble
Dutoit: Montreal SO (1920 version)
Rattle: Nash Ensemble
Salonen: London Sinfonietta (1947 version)
Stravinsky: Cologne RSO (recorded 1951) (1947 version)
Whitfield: Endymion Ensemble (1920 version) [nla?]

I did a reasonably blind listen without following the score, simply to see which versions appealed to me more than others.
So I didn't concern myself with adherence to tempo or dynamic indications, but I did have two specific criteria.
As Malcolm MacDonald says in his preface to the Boosey & Hawkes edition of both scores, Stravinsky went back to the word's [symphony] ancient connotation of groups of instruments sounding together, and used the plural to indicate that the music is made up of several of these instrumental colloquies. So I particularly listened to how the instruments blended (or not!).
The other was the 'sourness' of what I assume to be a fiendishly difficult 'motif' for the four horns (we need a forum expert to tell us just how hard this is to sound right!): in the 1947 score it appears as the two bars before figure 67, the third and fourth bars before figure 68, and the two bars before figure 69 (not identical in each case); the two bars at figure 43 for the four horns, bassoons, and double bassoon are similarly challenging.

What I discovered was that my preferred versions were those from smaller ensembles rather than orchestras, and I do wonder if this is because they are used to playing together in smaller groups, and perhaps more readily listen to each other and blend better.

So my personal 'favourites', which I intend to follow more closely with the score (only done once so far) are:
Nash Ensemble/Rattle (1947)
London Sinfonietta/Salonen (1947)
Endymion Ensemble/Whitfield (1920)

I am very interested to find out what other forumistas make of this work.
Over to you!

[Pulcinella]

Instrumentation

1920 version

3 flutes (third doubling piccolo)
Alto flute in G
2 oboes
English horn
2 clarinets in B flat (first doubling clarinet in A)
Alto clarinet in F
3 bassoons (third doubling double bassoon)
4 horns in F
2 trumpets in C
Trumpet in A
3 trombones
Tuba

1947 version

3 flutes
2 oboes
English horn
3 clarinets in B flat
3 bassoons (third doubling double bassoon)
4 horns in F
3 trumpets in B flat
3 trombones
Tuba

[Pulcinella]

Good choice, Pulci, for our first of this summer. And fine list!

At least, in this part of the empire, we don't have to complain about twofers, my fav being ignored, presenter bias etc etc.

Oh: I don't know about that; just wait till the discussion gets going.


Happy to amend the available recordings list if others can identify the version (or split it into 1920/1947/unidentified) if that helps.
Yes: a happy coincidence to do our own tribute to a tribute to a Centenary Proms Composer as the 2018 Proms season kicks off.

[Richard Barrett]

I haven't yet had a chance to do any comparative listening but I will say that I much prefer the 1920 version, from a time when Stravinsky was in general more inventive with timbres than he was interested in being in the more sharply-etched outlines of his neoclassical works. The 1920 version furthermore sounds more to me like the memorial to Debussy that it was intended as, whereas the 1947 version seems to have forgotten that aspect of the piece altogether.

One reason for preferring recordings made by chamber ensembles could be the recorded acoustic, which is likely to be less spacious than a room otherwise used for full orchestras, but at this stage I wouldn't go so far as to give a more definite opinion on that!

If I want to listen to it I would instinctively reach for the last Boulez recording. But I'll check out a few more as time permits in the coming days. Thanks for compiling all this information!

[Pulcinella]

I haven't yet had a chance to do any comparative listening but I will say that I much prefer the 1920 version, from a time when Stravinsky was in general more inventive with timbres than he was interested in being in the more sharply-etched outlines of his neoclassical works. The 1920 version furthermore sounds more to me like the memorial to Debussy that it was intended as, whereas the 1947 version seems to have forgotten that aspect of the piece altogether.

One reason for preferring recordings made by chamber ensembles could be the recorded acoustic, which is likely to be less spacious than a room otherwise used for full orchestras, but at this stage I wouldn't go so far as to give a more definite opinion on that!

If I want to listen to it I would instinctively reach for the last Boulez recording. But I'll check out a few more as time permits in the coming days. Thanks for compiling all this information!

I think that I prefer the 1920 version too.
Malcolm MacDonald, in his Preface to the published scores, says this, which I think sums up the difference perfectly:

Briefly stated, the original version is the more redolent of the liturgy and Russian orthodox church music; the revised version is more abstract, more a Cubist play of colours and planes.

[antongould]

Scary I've either never listened to this work before, or perhaps never listened to it properly. I find it quite wonderful - but not having access to pulcers 11 versions I have to restrict myself to the 3 on Amazon Prime Music - all 1920s and only one, the Craft 7:59 in the original list ..... Will my life be ruined without 1947 ..... ????

[Joseph K]

I think that I prefer the 1920 version too.

Make that three. I've just listened to the 1920 version, NYPO/Boulez for the first time. The version in which I've hitherto known this piece is the later one conducted by Stravinsky but I prefer these tempi and acoustics as well as the more plangent, affecting sonorities of the scoring.

[BBMmk2]

Starting a week too early Pulcie?

[Pulcinella]

Starting a week too early Pulcie?

Not really, Bbm: Alpie puts up details a week before the traditional BaLs to get the ball rolling.
Not sure we'll have come to a resolution by next Saturday though.

[ChrisBennell]

I have a recording from R3 of a Boulez / BBCSO concert given in 1989 at the "Huddersfield Contemporary Music Festival". Originally on cassette, but I transcribed it to CD some years ago. I'm very fond of this performance and recording - it takes 9:09 but I've no idea whether it is the 1920 or 1947 version. The concert also included Le Rossignol, which I also recorded. Not sure whether this orchestra and conductor ever recorded it.