As a result of this earlier discussion,
I've created an artificial audio file to illustate how the first movement
would sound if Tchaikovsky's metronome markings were followed. I was very
surprised to find that I have never heard the main theme played at
this speed, or even anywhere close, and my first thought was that I must
have misread the metronome marking of the Allegro con anima (q. = 104), but it's marked correctly in the first
edition and all subsequent editions I've come across. The effect of taking
this tempo literally is to completely change the character of the movement,
from the lugubrious funeral march we usually hear, to a brisk and
forward-thrusting main theme (from 1:54 in the recording, after the slow
The section that Yoon Jae Lee queried in the previous discussion comes at
4:27, marked "Molto più tranquillo (q. =
92)", and it too is actually much faster than the slow waltz that most
conductors (with the exception of Igor Markevtitch) offer instead.
The total playing time for the movement, taken strictly at Tchaikovsky's
tempo markings, is 11:47. It would be very interesting to know if there are
any recordings that are anywhere close to that speed.
With the usual disclaimers and apologies for the limitations of
artificial audio, you can listen to the audio file
There is a recording by Markevitch and London Symphony Orchestra (the
album include the six symphonies) that is close to Tchaikovksy's metronome
markings. The total playing for the total first movement by Markevitch is
My favourite performance of this symphony is Herbert von Karajan's
recording with the Berliner Philharmoniker, re-released by EMI in 2007.
Admittedly the first movement (15:51) is very much slower than your audio
file observing the correct metronome markings, but this is partly explained
by the fact that von Karajan takes a good 3 minutes over the slow
introduction. Under his direction the emotional impact of the main theme is
surprisingly similar to the ideal performance you have re-constructed. It
sounds powerful, thrusting, almost aggressive. There is nothing in the least
bit lugubrious about it, even when it fades into the distance at the close
of the movement. The mood there is one of grim determination, not of
neurotic despair. In fact von Karajan makes the entire symphony sound
positively heroic. He was notoriously cavalier about authenticity, disliking
repeats so much that he even failed to observe the one marked in the first
movement of Mendelssohn's Italian Symphony, completely distorting the
movement in the process. So metronome markings were probably of no interest
to him whatsoever. But he does seem to have penetrated to the heart of
Tchaikovsky's Fifth Symphony, allowing it to take its rightful place among
the so-called 'Great Fifths', by Beethoven, Bruckner, Mahler, and Nielsen.
(In my opinion it is a finer work than the fifth symphonies of Sibelius,
Prokofiev and Shostakovich.)
In my opinion there is a contradiction between the metronome markings and
the words used to indicate the speed. I will try to expain:
18 bars before the waltz section there is the indication "Un pochettino
piu animato. Thie indication has as reference q.=104. ¿how much is
pochettino?. I think that q.=110 is good. When, 18 bars later, the famous
waltz theme appears, it is marked as "Molto piu tranquillo" (q. 92). is
"Molto piu tranquillo" a good expresion to indicate the change from (q.=110)
to (q.=92)?. I Think not. The words "Molto piu tranquillo" seems to indicate
a speed lower than (q.=92).
In adittion to the mark "Molto piu tranquillo", the violins have the
indication "molto cantabile ed espr.". Molto cantabile ed espressivo and
speed (q.=92) is difficult to reconcile. The theme character, with the
syncopes, requires a speed lower than (q.=92) to play it "Molto cantabile ed
I recently heard a recording of the symphony by Daniele Gatti and the
Royal Philharmonic Orchestra. The first movement takes 12m 45s (as
Markevitch version) but Gatti plays the waltz theme at (q.=56) much slower
than (q.=92) and Markevitch.
But surely "tranquillo" reflects the mood of this passage, not the speed?
In any case we have a defintiie tempo marking of (q.=92) — repeated at the
same point in the recapitulation — which is routinely ignored by 90% of
conductors. There's absolutely no excuse for this, especially when they're
content to observe Tchaikovsky's metronome elsewhere in the work when it
Forgive me, I'm feeling rather tetchy after coincidentally hearing a new
recording of the symphony earlier today where after a good start the
conductor inexplicably slowed the pace to a crawl at exactly this point !!
My hypothesis is that (q.104) for the "allegro with anima" is correct but
(q.=92) for the waltz section (marked "Molto piu tranquillo") is wrong. This
passage played so fast (q.=92) makes nonsense, so, almost all conductors
ignore this indication (q.=92) and follow what "Molto piu tranquillo" and
"molto cantabile ed expr." suggest for each conductor. There are various
versions with the "allegro with anima" played at speeds close or equal to
(q.=104) but i only kown a conductor to play the waltz section at (q.=92),
There's absolutely no chance that Tchaikovsky made a mistake with the
metronome marking, since he checked the proofs of the published score, as
well as conducting from it, and the same marking appears in Sergey Taneyev's
piano duet arrangement, made in consultation with the composer.
Yes, it may sound strange when we're used to conductors stepping on the
brakes at this point, but it's not so long since conductors said the same
about Beethoven's "nonsensical" tempo markings (the opening movement of
his Fifth Symphony being a prime example!).
But in Sergey Taneyev's piano duet arrangement, only is the mark for the
'allegro con anima" (q.=104). For the waltz theme there is not mark, only
the indication "Molto piu tranquillo" and "molto cantabile ed espress." You
can check it here:
The pdf is here:
Sorry, Manuel, my memory must have been playing tricks on me.
Nevertheless, you can see from the Jurgenson edition available from the same
site that the metronome marking appeared in the very first edition of the
full score (in both occurrences of the same passage), and the Soviet
critical edition of the composer's works confirms that it was in the
autograph manuscript as well.
Mr. Langston's computer-generated realisation of the Fifth Symphony's
opening movement gives a valuable insight into how it might have been heard
in Tchaikovsky's own day. The setting of the first subject at 104 dotted
crotchets (or, in American terms, dotted quarter-notes) per minute makes it,
indeed, startlingly dynamic in comparison with most recorded performances.
With regard to the vast majority of well-known works in the concert
repertory, irrespective of composer, there is a regrettable tendency towards
an accretion of standard ways of doing this or that passage. Sometimes to
clear that aside is an instructive thing in artistic terms.
On a further general point, it is by no means unusual that
Romantic-period composers' metronome markings were set aside in
twentieth-century performances (and on into our own times) in favour of
slower tempi. These seem to result from an exaggerated focus upon the
lyricism of the music, to such a degree that entropy can even set in.
No less outstanding a conductor than Leonard Bernstein was capable of
misinterpreting Tchaikovsky's Sixth Symphony so deeply as to slow it down
virtually to a standstill.
The reason why I think the indication (q.=92) is wrong for whe waltz
passage, is because this mark is contradictory with the indication "Molto
piu tranquillo" in its speed context. How should play a conductor this
passage, at (q.=92) or "Molto piu tranquillo" ?. This is a dilemma for the
conductor because the indications are contradictory and most conductors
(actually all but Makevitch) follow "Molto piu tranquillo" and no (q.=92).
The indication "Molto piu tranquillo" is in all editions and arragements
(for piano solo, piano duet, two pianos, and orchestal partes) but the mark
(q.=92) is in some and no others (curiously in the orchestal parts, edited
by Breitkopf & Härtel
are no metronome markings).
The reason Tchaikovksy wrote both the mark (q.=92) and "Molto piu
tranquillo" is a mistery for me. A solution to this dilemma is that
Tchaikovksy wrote (q.=62) and a human error changed the 6 for a 9 in the
I can only restate the two points I made earlier:
- The metronome mark of q=92is not 'wrong'. It appears in the autograph
score and the published score by Jurgenson, for which Tchaikovsky checked
the proofs (and from which he also conducted the symphony).
- That "tranquillo" isn't referring to the tempo — it's an indication of
mood (i.e. less tautly than the previous passages).
This is a terrific forum. I will be conducting this in October hence my
metronome marking googling. I am torn between my (near) fanatical adherence
to the composers wishes and my desire to be re-engaged as guest conductor.
The evolution of tempo choices is inevitable. I find that generally it
seems the tempi of classical repertoire get faster as the years go by and
romantic get slower. As fascinating as the mp3 of the symphony was - I just
couldn't do it. I have the Markevich recording and even at his "near to
marked" tempi, I get uncomfortable.
Thanks for your help in this issue. It has helped me to get at least
closer to doing the right thing musically! Daniel Warren
To Mr. Warren: good luck with your forthcoming performance of the Fifth
Symphony! The composer's metronome markings are such a clear base for the
overall shaping of the work that the closer a rendering of it comes to
these, the better. The foreground and background elements of the score
emerge more clearly at the sequence of tempi that the composer has
designated, and one difficulty that comes with the slower tempi that
conductors have often chosen is a loss of momentum in the elaboration of the
opening movement's first subject from the upbeat to measure 66 onwards. No
orchestrator has a better sense of how to preserve a sense of flow in a
composition than Tchaikovsky, and slower speeds than those he preferred are
harder to employ and still produce a comparable result. I personally feel
you are entirely justified in preserving a concern for a composer's tempo
indications. All the best, in any case, with your performance.
I for one would love to hear Tchaikovsky`s 5th symphony, as in any other
of his compositions, played as he indicated, otherwise, Daniel Warren, why
would he bother doing so and why would I want to hear it? Otherwise its
something else entirely, and still perpetuates the `tortured Tchaikovsky`
image, we are trying to change? Please have the courage to perform at the
right tempo, otherwise we are stuck with this status quo, we need someone to
break the mould....
I don`t see why `the tempo evolution is inevitable`, to me it's more the
evolution of the conductors, and performers' own ego
Of course, indications of the author are important ... but all the
interpretive literature (especially since there are discs) is an eternal
debate about the thousands of possibilities (theoretical) dates to the
interpreter. Think of Toscanini, for example. The music is an
interpretation, even more than performance, I think.
How many books talk about it! For example (in my language...):
- Book Title: Music and interpretation. Subjectivity and knowledge in
- Authors: Sini Carlo Santi Piero Enrico Fubini
- Publisher: Trauben
- Series: Musicology
- Date of Publication: 2003
- Genre: Music
- Topic: Music
- Pages: 194
- Curator: L. Attademo
- ISBN: 888839821X 9788888398211
If by contradiction we take these thousands of possibilities, the music
would become a little boring ...
Listen to these two interpretations of the famous "Fifth"Symphony of
Tchaikovsky (final) ... what about?
At first, Tchaikovsky have consistent sense of tempo and in scores he
directed tempi quite precisely. Tchaikovsky used the term "tranquillo"
as antonym of "con anima" or "con spirito", and it reversing "piu mosso" or
"Spirito" and "anima" both mean "breathing" or "respiration".
Breathing makes motion and action, so "con anima" and "con spirito" mean
"lively". By using "tranquillo" he intended not only "calm" but also "a
little bit slowly".
[Symphony No.1 - 1st movement]
- "Allegro tranquillo (q. = 132*)" - "Poco piu animato"
*Tchaikovsky's "Allegro giusto" and "Allegro moderato" eaqual "q. = 132
[Piano Concerto No.1 - 3rd movement]
- "Allegro con fuoco" - "Molto piu mosso" - "Tempo primo ma tranquillo"
[Souvenir de Florence - 1st movement]
- "Allegro con spirito" - "Tranquillo" - "in Tempo giusto"
[Sleeping Beauty - #10 Ent'act et Scene de la chasse royale]
- "Allegro con spirito" - "Un poco piu tranquillo"
[Symphony No.5 - 1st movement]
- "Allegro con anima (dq = 104 )" - "Un pochettino piu animato" - "Molto
"Un pochettino piu animato" may be "dq = 108 to 112, so "dq.= 92" for
"Molto piu tranquillo" is definitely suitable.
If he had intended to change it more slowly, he should have used words as
"Molto meno mosso".
And it is important to be performed as rhythm of waltz. Performing slowly
it would not be a waltz that Tchaikovsky intended the same tempo as the
waltz of the 3rd movement.