I lost where this was being discussed a bit elsewhere so and am starting a new thread as I wished to add to the discussion.

It has been said that the accompaniment vs melody rubato -- what I am calling Musical Dissociative Disorder -- is something that is best exemplified in the vocal literature, as I presume, a demonstration of melodic freedom of the singer, etc. However, I would have to say that listening to a performance where the accompanist (pianist or conductor) did not in fact accomodate the accompaniment to the singer's interpretation, thereby maintaining the vertical integrity of the composition and limiting rubato to the tempo as a whole, would be looked upon as a poor accompanist indeed. I'm sure many pianists here have good or even extensive accompanying experience (I do), and the fact that no matter how flexible the soloist or conductor (choral works) can be, that the pianist can "follow" is recognized as the achievment of art and skill -- just plain ensemble ability. This all came back to me as I was listening to a Chopin Nocturne performed by violin (melody) and piano (accompaniment) and recognized as I listened that I would be horrified to hear any dissociation of the melodic rhythm from that of the accompaniment in this two-performer version. Why should it be any different if performed just by a pianist? I maintain that such a dissociation is both unmusical and contrary to everything a musician trains by.

Hi Eddy,

I too have accompanied singers and totally agree with your viewpoint. The pianist has to be flexible in order to follow, never lead, the singer. In following the singer, the pianist plays an almost imperceptible nanosecond (figuratively speaking) behind the singer. Too there are those unwritten subtle pauses to enable the singer to take breaths as the song unfolds.

Personally, I've never put much stock in those 18th and 19th century descriptions of rubato whereby the melody flows freely, accelerates then slows to resume the pace while the accompaniment has remained strictly in tempo and meter, as Chopin taught. It would lead to the very disassociation that you describe. In fact, the pianist ought to follow and mirror the singer through any rubato, ad libitum, or a piacere or other subtlety in my opinion. Likewise, in the piano literature, I would assert that it is not just the melody, but rather the whole musical fabric that is affected by rubato. I would argue that playing rubato in this way is more coherent, dramatic, satisfying and convincing, at least in my opinion.

David

David,
I appreciate your commenting on this subject. Earlier this evening the thought occured to me that, perhaps, this notion of free melody versus "fixed" accompaniment is one erroneously born in the auditor when listening to several of those passages in Chopin where irregular groupings of notes are accompanied by clearly patterned accompaniment, the impression being that the melody has "done what it liked" with the tempo when the accompaniment has not -- when of course all along, it was just an irregular grouping in the melody. Examples are almost everywhere in his writings, but two examples suffice to illustrate my point: the 4th measure of the "A" material of the "Raindrop" Prelude, Op.28, No. 15, both times but especially in the return of the A section; the 2nd and 3rd full measures of the first Nocturne, Op.9, No.1. I think that someone listening to these for the first time may have had the impression that Chopin's RH was freely singing while the LH was keeping careful time.

Having said all this, I can't deny that Karol Mikuli (his pupil and then teaching assistant) confusingly states the following:

I really want to give Mikuli the benefit of the doubt ... but struggle to do so. Perhaps all this he's just described was the RH in irregular groupings or fioritura passages? Oh the mystery.

Eddy

As I mentioned in the other discussion, this very possibility was mentioned while I was still learning the piano and was associated with Chopin only. The piece I was learning was the Prelude in d (24) The teacher studied at the Paris Conservatoire in the '40s, if this has any bearing on the discussion. To add to this, I never finished that particular prelude and never went any further into the matter and had quite forgotten about it till this came up. I also remember that I was told that Chopin was to be played in strict time and not with the rubato that so many pianists employ. One example of good playing of Chopin that was given to me was by Nelson Freire. We did discuss Arau, but if I remember correctly he was mentioned as paying correctly, which I understand to mean that he played what he saw on the page.

There's a difference between "accomodating the accompaniment to the singer's interpretation" and pedantically following every microscopic nuance of the singer. As a professional accompanist, I want to be sure that I can follow the singer (or instrumentalist) as closely as I want--but it doesn't always mean I should. There are times in music when a "soloist" (I hesitate to use that word, but there isn't a better) wants to be able to push against a firm rhythmic structure without it giving way. If the accompanist is too "sensitive", it can cause the performance as a whole to lack conviction.

To be fair, this sort of rubato is more common in popular music, jazz and musical theatre (especially where the composer has set speech rhythms in a reasonably natural way), and also in the virtuoso violin repertoire, than it is in mainstream lieder or art song.

Since you seem absolutely convinced that this sort of rubato is wrong, it's unlikely that I can persuade you otherwise just by typing a few words. But I can tell you firstly that top professionals occasionally do this deliberately, and secondly that you've probably heard it without being aware of it. (If it's obvious that the pianist isn't following, then they're overdoing it.)

Okay, here's the first two examples I found on YouTube.

Martha Argerich plays Chopin's Mazurka opus 24 no. 2. Pay attention to the section from 0:37 to 0:48 (where Chopin actually writes the word rubato in the score); it's subtle but it's definitely there.

Maria Callas sings the Habanera from Carmen. There are a few examples of rubato in this performance. I think the clearest is at 2:12: she wants to emphasise the text "Il est la", making the last two quavers of the bar late; but if the orchestra were to follow her, it would spoil the Habanera rhythm, so the orchestra keeps strict time here.

There would be plenty more examples out there (it didn't take me too long to find these two), but I'm not going to spend any more time identifying them for you. Just keep your ears and your mind open.

Alexander,
Thank you so much for the examples and I weigh most carefully what you offer. I listened to the Chopin VERY HARD, over and over, and I just can't appreciate any dissynchronization between the hands. Maybe if I could slow it down I might be able to detect it. I would love it if you could find something a little easier to appreciate -- a Nocturne perhaps? In fact, I don't think there would even be such a notion as this subject if it were this hard to hear.

Eddy

I think this is very likely the case. Mikuli's comment on this, and others' echoing of it, is an issue that has long confused me, for it it doesn't seem possible for melodies and harmonies ever not to match up with one another except when it is specifically part of the musical structure. Keeping in mind that playing the piano is largely an illusion anyway (in terms of the listener's appreciation of it), I'd say too that the likelihood is that Chopin was able to create the illusion that his left hand was keeping perfect time (and presumably his auditors not having a metronome handy to check him ), probably because his rubato was so perfectly and aptly applied. And indeed, the filigree figurations in nocturnes that are deliberately asynchronous help create that illusion anyway. I think it is really almost a logical impossibility for the two to diverge and sound like anything other than cacophony.

Is it such an absudity that one hand is precisely in time while the other deviates only very slightly? It is not that one is to have the left paying bar 10 while the right plays bar 12! I ask you, if you have this pattern: 5 notes against 6 or 7 against 11, how can you possibly play these if not by taking these infinitesimal liberties? I realise this is not quite rubato, but it does come near. Maybe we should be discussing the amount of rubato.

No, in fact it is not rubato at all. This is simply polyrhythms of one part against another, both of which can be played exactly in time if one wishes (and can hear the polyrhythms against one another). It's probably easier to visualize by thinking of a 2 against 3, which is no different an idea in concept but much easier to count out. Both parts in a two against three are often played exactly in time. Rubato, on the other hand, is literally the idea of stealing time from your overall tempo, then in many cases presumably making it up so that the effect is natural, though the latter is not necessarily a requirement.

Indeed, though 2 against 3 is equal to 4 against 6, which is perfectly divisible.

Anyway, I have never played the way Eddy is protesting against, so...

Very true, though it might be harder in actuality for the ear to hear a 4 against 6 while playing.

In your initial response, though, you did say 5 against 6, though, which is a different kettle of fish entirely In the opening of the B-flat minor nocturne, op. 9 No. 1, I believe there are some weird polyrhythms of like 11 against 6. In such cases, I think it's permissible to break up the righthand figurations by, say, 2-2-2-2-2-1, and when it's played up to tempo no one will notice anyway.

Interesting discussion, guys! I was going to add my two cents but keep deleting everything. Too hard to describe my thoughts on rubato, except that it's hard for me to do! Maybe because I'm a stickler for sticking with the rhythm that's on the page. Also, I don't care for players who push and pull things so much - it gets annoying real fast.

Anyway, carry on.....

I was thinking of this as practice and a sure way of getting 2 against 3 right.



I suppose I was thinkning of that as a type of rubato and I wonder if this might not be what Chopin is said to have done. I believe that unless one is an African drummer 5 against 6 will never divide evenly.

Indeed, Monica. Maybe you might listen to some of the Chopin performances on the site. I listened to one of the Prelude in e... I do not believe there were two notes for the left hand that had the same values.