Hand crossing in Liszt Sospiro

[diomedes]

Thanks to our friend Marc Andre Hamelin, I made a valuable discovery, I'd like to pass on the advice.
Bar 20 the octave E and A in top stave usually played by left hand, the rather impractical hand crossing can be made tolerable by taking the first note with the right hand, so left hand when it crosses for the e and a plays a single note.

Wish I noticed this before.

[perfect_pitch]

Why the need for a post simply to state that another hand may play the note the other hand is supposed to play???

Seriously??? I've played this piece before and it's perfectly playable exactly as Liszt notated it.

[diomedes]

It's possible that someone may find it useful.

It's also possible that you're a better pianist than us all including Hamelin. Anyway, accept my apologies for upsetting you~ 

[stevensk]

Why the need for a post simply to state that another hand may play the note the other hand is supposed to play???

Seriously??? I've played this piece before and it's perfectly playable exactly as Liszt notated it.

-Calm down perfect pitch, New ideas are interesting.
In this case I dont agree whith diomedes suggestion/solution but I like outside box ideas 

[cbreemer]

Even when something is "perfectly playable" as written there is no harm in letting one hand help another. I find myself doing that a lot even in simple stuff like Kinderszenen. I find it heartening that even a Hamelin seeks to make life easier whenever possible

[pianoworthy]

Well Hamelin is well known for taking liberties when it comes to fingering. He's a genius at coming up with fingering solutions that are outside of the box.

As long as the music is still effective, I don't see the problem.

[perfect_pitch]

Have you thought that there may have been a deliberate reason that Liszt didn't write it that way???

After all, Liszt was a genius, so he could have possibly entertained the fact that it could have been shared with the hands, or he deliberately did it to secure the arpeggiation remains unbroken and uniformed in touch and sound by playing it with the LH.

Granted Hamelin is brilliant, but so was Liszt.

[j_menz]

Have you thought that there may have been a deliberate reason that Liszt didn't write it that way???

What way? I don't see any fingering indicated. Seems to me Liszt just wrote the notes, without instruction.

[diomedes]

Have you thought that there may have been a deliberate reason that Liszt didn't write it that way???

Absolutely. There's a reason for everything, and as a principal serious thought should be given to it. Just like there's a reason for your question mark enthusiasm, you must really enjoy question marks a great deal.

But in all seriousness, a score being provided by a composer is a clear form of communication, so without a doubt, serious thought goes into what's there and what it means. But what if the composers attempt at communication is unclear. This happens very often, and not necessarily their fault. And also, a performer has a responsibility to decide what to convey. Is an overall concept or clarity more critical than a small detail.

Now, I continued to inquire into this and dug up interesting things. When Hamelin crosses hands, it looks like he uses a long figer for the left hand when it crosses. It appears like it's one motion, fluent and certainly he doesn't use the thumb. In hindsight, only an idiot would use a thumb in that instance, like me....:

I learned this from a henle edition, which is ok i guess. I looked at Hamelin again, it looked like he uses 5th for the lower note and 2 for the octave higher one. I tried it. Incredible. Accuracy up about 50%. No more scrambling awkwardly for the bass note. Then i went to imslp, there was an edition by Emil Sauer, and it cites to use 3 then 2 for the upper note. Well, who on earth would use 3 and 2 for an octave, but in this context it increases the accuracy even more, and the gesture is incredibly economical. Looks like left hand plays one note, so the distribution wasn't outlandish, but the proper way of doing it is very much o outlandish. And it works, who knew.

When I was doing the Alkan 30 ans at the end of the development was the hand crossings of blood and murder, i didn't give a single thought to the distribution. The way it was was how i did it. It caused wrong notes, and i think in that extreme case it's ok, even though i hate wrong notes more than anything. Hamelin redistributes that part, in effect simplifying it. I also remember some guy on youtube telling me to redistribute it after seeing my video, but my reply to that was if there was a way to simplify the opening of the Liszt 1st concerto by redistributing it, would you? Pretty sure no one would say yes.

Anyway there goes that mystery. And now, for the next controversy, i'll continue working on the tempest and debate with my cat if i should use hand crossings in there also.

[j_menz]

if there was a way to simplify the opening of the Liszt 1st concerto by redistributing it, would you? Pretty sure no one would say yes.

Why not?

And redistribute from what? Liszt's instructions (where I might think twice) or some poncy editor (where I wouldn't)?

Aren't you confusing some old habit for holy writ?

[diomedes]

It's a purely hypothetical question, obviously you can't. And, the reason I would not is due to the intuitive gesture of that opening being physical and theatrical.

Of course, at this point it's points of view based on an individual's disposition.

[michael_c]

Have you thought that there may have been a deliberate reason that Liszt didn't write it that way???

The way Liszt wrote it is an indication of how he wanted it to sound. It's up to each pianist to decrypt this indication, decide how they think it should sound and find the best way for them to create the sound they aim for.

The proof of the pudding is in the eating and the proof of the pianist is in the hearing. As far as I am concerned, a pianist may redistribute notes between hands as much as they like, as long as it sounds right.

[gustaaavo]

I've always thought that those unorthodox fingerings were actually Liszt's own (because of their weirdness and because they appear in most editions). In fact, in the book "Conversations with Arrau", Arrau makes the point that those weird fingerings in Liszt should be taken seriously, and that he had learned a lot about the piano from them. Does any of you know if those fingerigs (or which of them) are from Liszt?

Nevertheless, my teacher used to tell me (and now I agree with him) that Liszt's (and Arrau's) technique is simply different to his or mine, so that there's no point in suffering (e.g. in Gnomenreigen there are fast ascending chromatic thirds with 4-2 in every one of them) just to play with the same fingering as Liszt. What is important, as has been pointed out, is to make it sound right.

So now to the measures mentioned in OP. I used to play the bass melody (bars 22-24, notated for rh only) alternating hands (E with left, f# right, A left...) and distributing the arpeggios also between them (if I'm not clear and someone's interested I could scan my score). What is important here is to have smooth arpeggios (and if you can play the opening you can distribute arpeggios smoothly between the hands) and to bring out the melody. So to play it as notated does not seem at all necessary and is a lot harder.

On the other hand, for the octave A in the second beat of measure 22, I think your suggestion does affect the result negatively. The octave A should sound different from the C#, which is part of the arpeggio. So, unless you're a genius at voicing, you are playing A as the top note of the arpeggio and resolving the octave B to a sixth (C# and A).

Edit: I just noticed OP was about measure 20, not 22 (the image had markings in 22, not in 20). I won't delete what I wrote as it may be useful anyway. But now I'll run to the piano to try your suggestion.

[diomedes]

Bar 23 I distribute in favor of economy of motion, I don't hear any sound issues. Maybe I'll post my recording of this when it gets done.

As for the 3,2 with left hand crossing over, the fingerings are not in italics, with what I have impossible to tell. But having tried it for a day, I think Liszts understanding of the piano would have caused him to take the same approach.

[diomedes]

Discovery of discoveries:

I found a Lang Lang performance (both visual and aural) that did not disgust me, despite very significant liberties. Look it up on youtube. He does justice to the poetics of the musical language unique to this music, but.... adds octaves frequently and does away with the hand crossings entirely, so the broken octaves are played by the right hand alone. I'll comment that the hand crossings are a very significant technical issue, especially when it moves to E major. I understand that at his tempo, it's the only feasible way of getting the tempo and accuracy right.

I think this is the first video of Lang Lang playing that i respect, and it takes big liberties in places. And it doesn't have the deranged faces of his usual performances that, in my opinion, translate into sound and effect the music. Worth viewing if it's news to you.

[vansh]

Looking at the music, on a similar vein, what about changing the hands for the arpeggios? I noticed that in a lot of places, the right hand is just playing 3 notes of the arpeggio, when it seems like it would be easier to have the right hand play 4 to make it easier on the left hand (since it's the left hand that's moving far away and thus needs more time). For example, in measure 10, as written, the left hand plays:

low Bb (over an octave below the Db)
Db
F
Bb

while the right hand simply plays:
Db
F
Bb

If the right hand plays that Bb, then the left hand fingering could be 5-2-1, but if the right hand plays as written, then the left hand requires a bigger jump (so something like 5-3-2-1). Wouldn't it be easier then to play 4 notes with the right hand? Why is it written with just 3 notes for the right hand?

(Picture is of the measure in question.)

[vansh]

By the way, looking at a video of Hamelin playing the piece here:

For the passage in question (measure 20), he actually plays the E and A using his left hand's 5 and 2; his right hand doesn't help there.

[diomedes]

Looking at the music, on a similar vein, what about changing the hands for the arpeggios? I noticed that in a lot of places, the right hand is just playing 3 notes of the arpeggio, when it seems like it would be easier to have the right hand play 4 to make it easier on the left hand (since it's the left hand that's moving far away and thus needs more time)

I think that would help with bar 19 a lot actually. Something to consider when i return to it.

[vansh]

In looking at several recordings closely, I noticed that it seems like *nobody* really plays it as written.

What I mean is that in the broken octaves section (measures 13-20), as written, the melody is in triplets. The arpeggios are in septuplets (7 notes per quarter note). Thus, when the left hand plays the melody, the upper note of the octave should be played just after the right hand plays its second note on the descending part of the arpeggios.

What I see instead is that most pianists (including Hamelin and Lisitsa) will delay playing the right hand, such that the upper octave note by the left hand is instead played before the right hand plays the first note of those descending arpeggios, or between the first and second notes. This gives them enough time for the left hand to move back down and play the arpeggio notes after playing the melody notes. Which notes are played are as written, though (they do not redistribute the notes between hands during that section).

Lang Lang's rhythm is more accurate, but of course he does this by not switching hands at all -- his right hand plays the melody all the way through here, leaving the left hand to do the arpeggios. By consequence, although the melody is closer to triplets, the arpeggios aren't even either in timing or in volume.

I'm not exactly sure what to make of this. On one hand, I can only tell because I'm looking at them playing at 1/4 speed, while I can't really make it out at regular speed, so one can argue that it doesn't really matter if people can't really hear the difference. (Similarly, also everyone is uneven in the left hand when listening slowly to the 4:3 parts of Chopin's Fantaisie-Impromptu.) On the other hand, if people are arguing that it should be played as written, well, it doesn't seem like professional pianists do so, in order to play it at performance speed and have enough time for the left hand.

My thoughts are:
1. Switching hands is not only part of the technique of this piece, but also part of the visual flair. I'm surprised Lang Lang "unswitched" the melody because I would have expected him to be able to make the most of the theatrics with switching hands. (Instead, his theatrics are mostly with over-rubato-ing the phrasing.)
2. I'm a bit more undecided about redistributing notes in the arpeggios. It would be nice to play it as written, of course, but giving one more note to the right hand would make it much more feasible to keep the melody as triplets while playing the arpeggios with the correct timing. As for why Liszt wrote it that way (where the right hand isn't doing much when it could have played more notes, leaving the left hand to do most of the work), it could be as simple as he explicitly wanted people to work on fast left hand leaps (similar to La Campanella), or something more obscure as the publisher not wanting to typeset all those ledger lines below the treble clef when there's a perfectly good bass clef underneath it. What do people think?

P.S. Does anyone know what are Liszt's instructions regarding playing this piece? I've heard that the 3-2 fingering for the left hand for the melody in this section is part of Liszt's instructions, but I can't find anything about what else he wrote regarding this piece.

Hamelin's performance:

Lisitsa's performance:

Lang Lang's performance:

MIDI (to hear the timing "as written" so to speak):