In the opening bars of the last movement of Tchaikovsky's 6th, the strings have some four-part writing with a melody "F# E D C# B C#." However, the score reveals that the first and second violins are actually crossing voices at every note change, so that they alternate playing each note of the melody.

What purpose does this voicing serve? Is there a reason Tchaikovsky opted to do this instead of letting the first violin play the entire melody and having the seconds play "B A# G# E# E# E"?

enter image description here

  • The complete score can be found on IMSLP.
    – Aaron
    Commented Aug 30, 2020 at 6:42
  • 2
    A comment below a now-deleted answer mentioned portamento. If some of the slurs are played with portamento, the unconventional voicing could definitely have an effect. Judging from early recordings, portamento would have been much more prevalent in orchestral playing in those days than now. Can someone with more expertise than me in string playing comment on whether this could be a genuine effect? Are there recordings of this passage where one can perceive a significant portamento effect? Commented Aug 31, 2020 at 10:15
  • 1
    @JamesMartin In the Wiener Phil. video one can hear a slide between the first two notes of the Violin I part (B to E). However, the illusion (that the melody is being played by one section) is not as strong, since the Violin I part is quite audible. Even in this Wiener Phil. video, the violinists are not all using the same fingering (and are thus not all doing the same portamento) so any portamento effect would be diluted. I am not sure how a coordinated portamento would interact with the melody illusion.
    – angryavian
    Commented Aug 31, 2020 at 16:56
  • Note that also the viola and cello parts are written in a similar way. Thus the viola and cello take turns in playing the bass note. Commented Aug 17, 2021 at 15:05

4 Answers 4


In the 19th century it was standard to have the first and second violins on opposite sides (i.e. the second violins would sit where the cellos now normally sit). This kind of voicing would give a sort of stereo panning effect.

The Seoul Philharmonic Orchestra in the linked video has the second violins set up in the 19th century manner, but the effect is not really audible, possibly due to the way the concert was recorded and mixed. In a live concert situation it would probably work better.

  • 1
    To supplement this, at rehearsal mark G in the IMSLP score linked above, the passage in question returns, but in the voicing posed in the OP. Listening in the linked video to the opening (t=2240) and rehearsal G (t=2624), I don't hear an obvious difference. But in a live performance, given the "panning effect", section G would then sound like a unification.
    – Aaron
    Commented Aug 30, 2020 at 8:28
  • 3
    @Aaron It might just be the way it's recorded, but the stereo separation in that video is minimal, so the effect is not noticeable. But it's worth noting that the second violins are set up as Tchaikovsky intended.
    – PiedPiper
    Commented Aug 30, 2020 at 8:38
  • 3
    To add to this Musical Illusions Perfect Pitch and Other Curiosities with Diana Deutsch-- To Be Musical, Diana Deutsch discusses the left / right panning effect, but her conclusion is that even with left and right angular lines, we still hear the notes as if they are in conjunct motion.
    – user70304
    Commented Aug 30, 2020 at 9:40
  • @PiedPiper The stereo panning effect does seem like the only plausible reason. Do you have a source/justification for this claim (just want to clarify if this is your speculation or if you are referring to something), and/or other notable examples of this practice in other classical music?
    – angryavian
    Commented Aug 30, 2020 at 17:11
  • 1
    @OldBrixtonian I appreciated your answer and the reasons you gave, sorry I did not mean to exclude speculation.
    – angryavian
    Commented Aug 31, 2020 at 0:24

I discuss this conundrum in my book 'Musical Illusions and Phantom Words' (Oxford University Press, 2019). Apparently Artur Nikisch tried to persuade Tchaikovsky to amend his scoring. There's no question that the passage produces an illusion - at least for righthanders. I experienced it strongly when NOVA came to film my lab, and the UCSD Symphony played the passage as scored by Tchaikovsky, with the orchestra arranged in 19th Century fashion. One possibility is that the argument resulted from perceptual disagreements. If Tchaikovsky was lefthanded, that might account for it. There are pronounced differences between righthanders and lefthanders, taken statistically, in how the Scale Illusion - a stereo illusion that has a similar effect - is perceived.

  • 7
    Wow, thank you Prof. Deutsch for stopping by just to answer this specific question! Owain Evans linked a video of you discussing the Scale Illusion in another answer, and I just found the relevant excerpt (and video) from your book on Google Books which is very informative from a historical standpoint and also gives other examples of this illusion in classical music. It is unfortunate that we don't have a definitive answer for why Tchaikovsky wrote it like this, but it is great to read about what clues exist from you and other scholars, as well as about the auditory illusion it creates.
    – angryavian
    Commented Aug 31, 2020 at 4:14
  • 1
    Honestly I don't understand what this answer is trying to say. So Tchaikovsky was possibly lefthanded, and therefore would have heard the passage differently, and therefore... ?? Commented Aug 31, 2020 at 20:07
  • 7
    In listening to the Scale Illusion, which I discovered before I heard about this Tchaikovsky passage, righthanders tend strongly to hear the (illusory) higher melody on the right and lower melody on the left, regardless of where the tones are coming from. However, lefthanders are more likely to obtain complex percepts. The statistics from my original experiment are published in Deutsch, D. 'Two-channel listening to musical scales', Journal of the Acoustical Society of America 1975, vol 57, pp.1156-1160. The PDF is posted at deutsch.ucsd.edu/psychology/pages.php?i=107. Commented Sep 1, 2020 at 2:50

For reference, here's the comparison of the crossed and non-crossed version of the theme:

%%score [{T1 T2} A B]
V:T1           clef=treble
V:T2           clef=treble
V:A            clef=alto
V:B            clef=tenor
% 1
[V:T1] !tenuto!B2 (e^G)  (c3/2^E/2)  | c4   r2 || !tenuto!f2 !tenuto!e!tenuto!d     (!tenuto!c3/2!tenuto!B/2)   | c4 r2
[V:T2] !tenuto!f2 (^Ad)  (^E3/2B/2)  | E4   r2 || !tenuto!B2 !tenuto!^A!tenuto!^G  (!tenuto!^E3/2^!tenuto!E/2)  | E4 r2
[V:A]  !tenuto!D2 (FB,)  (B,3/2G,/2) | ^A,4 r2 || !tenuto!^G2 !tenuto!F!tenuto!^E   (!tenuto!B,3/2!tenuto!B,/2)  | ^A,4 r2
[V:B]  !tenuto!^G2 (C^E) (^G,3/2B,/2) | F,4  r2 || !tenuto!D2 !tenuto!C!tenuto!B,  (!tenuto!^G,3/2!tenuto!=G,/2) | F,4

There are multiple plausible reasons why Tchaikovsky chose to use the crossed version for the beginning.

  • As said by PiedPiper, Tchaikovsky would have assumed opposite-violins placement, which does give such interchanges a notable stereo effect. Even with an artificially hard-panned digital rendition over headphones it seems surprisingly hard to hear what's actually going on, however the non-crossing version does in comparison come over kind of flat and uninspired.
  • One reason for this impression is likely that the voices are very parallel without this trick. Now, in a way that's definitely “mission accomplished” – it very much is an emotional descent. And Tchaikovsky was no stranger to prominently parallel lines, the most obvious example being the Serenade for Strings – but here it's quite a bit more extreme, in particular with three consecutive sevenths between viola and 1st violins together with fourths between cellos and violas. The crossing version doesn't remove that effect, but it brings it over more sophisticatedly than if every single instrument is literally moving downwards almost all the way.
  • Within the individual voices, the resolutions are rather unsatisfying, in particular the E♯ in the 2nd violins would really rather resolve upwards. Well, in Tchaikovsky's version it does, albeit by a large jump.
  • Without crossing, none of the accompaniment voices really have any movement between the dotted quaver and semiquaver. That means the legato slur is instead a pair of tied tenuto notes – still a single bow stroke, but it doesn't have quite the legato quality anymore or alternatively doesn't separate the notes at all.
    Tchaikovsky definitely seems to have considered that (or maybe it was done later in typesetting), since the preceding notes are also tenuto and not slurred at all in the non-crossed version.
  • Some people like to see cross symbolism as a death metaphor all over the Patétique.
  • 1
    Thanks for pointing out the difference in bowing and articulation, that might be a big clue into Tchaikovsky's motivations.
    – angryavian
    Commented Aug 31, 2020 at 0:44

I think the string crossings encourage the players to make each note extremely expressive. You can see this especially in the 2nd violin part, with its jumps from F# to A# to D natural to E# to B. It would take a lot of browbeating from the conductor to get the strings to play a simple downward scale with such intensity. And the viola cello/bass crossings create a thicker sound.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.