The same movement: a corner of William Walton’s Bagatelle #1
Prompted by a random end-of-year question from Mark Delpriora, which clearly got under my skin…
L’istesso mov. The same movement, the same speed. A tempo indication (in parentheses) at bar 65 of Walton’s first Bagatelle for guitar., where the texture changes from a driven 16th-note climax to a lazily floating melody supported by a Satie-esque harmonic oscillation. No sign of a ritardando indication: in fact no indicated tempo changes since the initial Allegro ♩ = 126c.
Walton’s intentions seem clear at this point. Don’t change the tempo. There’s no explicit rit in the two measures prior to this, though one might make some allowance for what seems like a natural expressive easing of the pulse to go along with the (explicit) diminuendo. Nevertheless l’istesso movimento is, as musical indications go, as unequivocal as it gets.
There is not a single recording (as far as my researches with the help of Spotify go…) that honours this. A couple of the first generation of performers/recordings get tantalisingly close (more later, though in general I won’t be naming names), but recent recordings from younger generations of players invariably slow down in the preceding measures (to a greater or lesser extent) and end up with a pulse anything between 75–50% or less of the original tempo.
Before spilling the beans, we need to ask about the provenance of this specific tempo indication. It’s a particularly important question for guitarists— so many pieces in the repertoire are presented to us through the prism of ‘edited by’, and the Walton Bagatelles are no exception. The composer’s manuscript that now appears to be reasonable well-known famously presents a version of the 4th piece in a different key from Bream’s edition (in B rather than D major, and with a tremolando figuration rather than harmonics in the inner part:
(It’s interesting and maybe significant that when Walton made an orchestral version of the bagatelles as Varii Capricci, he preserved the key of Bream’s revision, and not his original.) There are other (though less significant) divergences, but if we check the measures of interest in the first bagatelle, we see the same indication:
In Walton’s orchestral score, this is replaced with poco a poco meno mosso e rubato. But there’s no notated ritardando: even here the intention seems to be that the passage starts at the tempo in which the previous section ended. (Having said which, the only orchestral recording of the work currently available also slows — if a little — before this point and resumes at a slower tempo…).
So, how do the recordings measure up? The first point to make is that most take the piece a little slower than Walton’s c.126 quarter-notes per minute. 116–120 seems to be the general range, but this is not unexpected: a composer’s subjective impression of tempo often needs to be tempered by acoustics of an instrument, recording and performance space, and I’ll contend that this tempo is definitely within the bounds of circa. Secondly, and given whatever authority we want to grant to the first recording of the performer for whom the work was written: Bream’s 1973 recording (Julian Bream ’70s) eases into the section with a slight ritardando, but continues at around 85% of the performance’s original tempo (100 vs 118). However, when he recorded the bagatelles again in 1984 (in the album Dedication) Bream had already moved away from l’istesso mov: here after a slight ritardando in the two preceding measures, there’s an abrupt change of gear, with a tempo at c. 84 (some 70% of the preceding pulse).
The closest approach to Walton’s indication is achieved by Carlos Bonell, in his 1994 recording. Although the preceding measures slow a little more, the new material kicks off at around 95% of the preceding pulse. I find this effective, although Carlos’ rubato at the end of each of the measures disturbs the flow a little. (Disclaimer — I studied with Carlos in the late 70s/early 80s, and I’m sure I worked on the Bagatelles with him, as I was playing them in recitals at the time… happy days!)
A small sampling of eight more recent recordings. Tempo assessments are within the tolerances of the tap metronome on my iPad, and the rit? attempts a judgement on a scale of 0 (no rit) to 5 (the music stops completely) of the ritardando in the previous bars.:
Your mileage may vary, but I find some of these impossible to listen to. I ask myself why.
Partly, of course, it’s familiarity. I heard and played these pieces more-or-less as soon as they were published, the first Bream recording is relatively faithful, and it’s the version that’s stuck in my ears. But I can’t help feeling that the transition from the opening texture to the middle section in the same movement — l’istesso tempo — is a fundamental aspect of the meaning of these measures and of the piece as a whole. Composers do not write these indications without thinking about them — it would have been perfectly simple, and indeed natural, for Walton to write ritardando in the preceding measures, and meno mosso at m.65, and without exception all recent recordings behave as if that’s exactly what he did. But he didn’t.
Having been prompted by Mark into these investigations, and digging back into my own marked-up score, I find the following:
A fierce admonition not to slow down — necessary because of course the natural thing to do at this point is to slow. How do we make sense of the transition to this new texture without doing this? Remembering that Walton was an accomplished composer of film music, let’s think of the different kinds of cut that movie directors use. At this point we don’t need to crossfade (a ritardando into the new slower tempo) or fade out (bringing the music to a complete halt before continuing)— in tempo at least — but we can cut directly. What changes? Well, sixteenths give way to triplets, repeated notes are replaced by an open and more harmonic texture, and a rhythmic accent on the second beat of the measure is replaced by a melodic accent on the third beat. But most significantly, this new music moves within a metre of one beat per measure, not the strongly accented three of the previous music. Although clearly faster than many performers currently hear this, there seems to me to be a particular atmosphere — of insouciance, wistfulness? — at this tempo that’s entirely characteristic of the composer, and that’s completely lost by taking the music at a significantly slower tempo. (Additionally, the faster tempo prefigures the one-in-a-measure tempo of the second Bagatelle, where the nod to Satie’s Gymnopédies becomes an explicit acknowledgement.)
Why does this matter?
I want to take a composer’s intentions seriously. It’s complicated with our repertoire as guitarists, as so much of our recent repertoire arises from creative collaboration with non-guitarist composers, but paradoxically it’s the aspects of a score (for example, dynamics, tempo, articulations) which don’t pertain to the mechanics of the instrument (chord voicings, sustaining notes, playing certain notes as harmonics and so on) which, if they genuinely express a composer’s intent as to the meaning of a piece, are surely the simplest and most obvious to respect.
Secondly, if every performance and recording of a work ignores these things, we’re establishing a performance tradition, and an expectation that a particular piece sounds like this. I’m frankly alarmed at not being able to find a 21st century recording of this piece that gets anywhere near what Walton made an effort to communicate with that simple indication — l’istesso mov. I can’t believe that every guitarist who has recorded it has tried it at or near Walton’s indicated pulse and then rationally decided to play it at more-or-less the same — slower — tempo that now seems to be standard. There is a sense in which if all you hear is a recent recording or performance, you are hearing a different piece, one in which a unique and particular expressive character (that Walton-esque wistfulness) has been lost.
With thanks to Mark Delpriora for the original sand in the oyster, and Forbes Henderson for conversation and more.
Postscript 1
There’s some useful background to the Bagatelles — musical and anecdotal- in Kenneth Kam’s pair of articles for the GFA’s Soundboard journal (45.4, 46.1). Information on Kenneth’s research here, the papers are also available on academia.edu.
Postscript 2
Posting this on Facebook let to an interesting and fruitful set of exchanges, copied here with permission. Initial comments in bold, with responses beneath
Gerard Cousins
A very enjoyable and thought provoking thing indeed! -Happy New Year
· Reply · 6 d
David Harvey
And to you, Gerard! Thanks!
· Reply · 6 d
Gerard Cousins
I had to go exploring myself and found this interesting take https://open.spotify.com/album/4EsrdEFGKXDMAhbfJ4S0gy
· Reply · 6 d
David Harvey
Will check that out. There’s Walton’s own orchestral version, of course, and also a piano version by Ian Farrington: https://www.iainfarrington.com/walton-five-bagatelles.html
Walton Five Bagatelles
IAINFARRINGTON.COM
Walton Five Bagatelles
Walton Five Bagatelles
· Reply · Remove Preview · 6 d
Gerard Cousins
David Harvey fascinating — lovely to hear it on different instruments and see how the differing musical/technical difficulties are tackled. To go back to your point — when playing guitar it’s all too easy to get ‘carried away’ with making ‘magic moments’ (which cause a lot of ‘musical’ compromises/differences with the score) but I have to admit — I would not want to play/listen to the guitar if those ‘moments’ didn’t exist! Can we satisfy both? I’m not so sure…..
· Reply · 6 d
David Harvey
Turning on a sixpence at that particular moment from driven 3-in-a-bar to wistful one-in-a-bar at the same crochet pulse is definitely a magic moment for me. I think this is what Walton meant at this point. I worry about an expressive orthodoxy that has a smaller repertoire of choices!
· Reply · 6
Chris Susans
Great observations David. I suppose one might ask the question “So who’s right?” Once the final double barline has been written the child leaves home and has to survive in the big wide world! Please don’t answer the question!
· Reply · 6 d
Gerard Cousins
I’m sat here with the score (without guitar) and beginning to hear what you’re suggesting and it’s indeed very nice. Is it just a question of fatigue that players slow down here and take a well earned breather!
· Reply · 6 d
Carlos Bonell
Very interesting dear David. Lots of well-written insight, open-minded, not dogmatic!
· Reply · 6 d
David Harvey
Thanks Carlos Bonell! It was good to have an excuse to listen to your fine recording of the Bagatelles again! Happy New Year to you!
· Reply · 6 d
Jonathan Leathwood
Thank you for writing this, David, and making the case. I totally agree that the movement is absolutely wonderful if the tempo is maintained: as so often, you see that facsimile of the manuscript without any fingering and the scales fall from your eyes — Of course, it doesn’t slow down, is what you think when you see it. It’s a lovely middle section that seems intended to flow with the same momentum as the rest.
But there is a problem. If you play all those very high artificial harmonics at q = 126, don’t they just sound choppy and unmusical, and inevitably so? I’m not sure I’d want to get into a debate about whether it’s possible to make them work if only one can play the guitar well enough: to me, there should be an ease here that doesn’t require transcendental skill. For this sole reason, I’ve never been able to insist to my students that they maintain the tempo.
It seems uncontroversial to me to say that the idea of having the harmonics was premised on the idea of slowing down. Suggesting to a composer that a certain concept can work if they will accept a modification to the character is part of the true dark arts of performer-composer collaboration, one little discussed in the literature. To me, the moment the idea of using the harmonics came up, Bream should have said to Walton, ‘It will mean changing the tempo of this section: can we get rid of the indication not to slow down? Or is that non-negotiable?’
· Reply · 6 d
Steve Goss
Jonathan Leathwood I wonder who suggested the harmonics in the first place? Bream? Malcolm Arnold?
· Reply · 6 d
David Harvey
I think there’s an accommodation to be made (and it nearly always is!) with the q = 126 — there’s a lot in the movement that really doesn’t speak at that tempo. I also think it’s possible to ease off a little coming into the section to maintain at least the sense of ‘l’istesso mov.’, as I think Bream does in his original recording, and as Carlos does in his. But so many of the current recordings slam the brakes on at this point, which feels to me both unthinking and trite.
· Reply · 6 d
David Harvey
The harmonics are in the manuscript (at least this one). Without knowing more about when and how it came about it is of course impossible to say what state of collaboration between Walton and Bream it reflects (and Forbes also tells me that Malcom Arnold stayed with the Waltons for some of the time at which the Bagatelles were composed, his influence could certainly be there)
· Reply · 6 d
Jonathan Leathwood
Hi David: good to know that the harmonics came in at an early stage. Not having a copy of the manuscript, I rather assumed they were Bream’s suggestion. I do understand your big point: if you slow down only as much as needed for practical purposes, you are not obeying the literal instruction, but you do still get something that sounds characteristically Walton; but there are many performances that slow down far more than that, and while the results are beautiful in a local sense, it’s to the detriment of the movement, the balance of the whole set, and Walton’s voice as a composer. I guess I was just making the narrow point that ‘L’istesso tempo’ seems such a strong prescription — not meno mosso, which is the best we can do — that it adds an enormous amount of confusion. To me, ‘maintaining at least the sense of l’istesso mov.’ is a bit of a contradiction in terms: I mean, it would never have occurred to me to characterize Bream’s recording as respecting it.
But your essay helps me to understand what’s at stake and the importance of using all artistic means possible to fool the listener into thinking that the tempo hardly changed. I probably read your essay too quickly and got hung up on the literal meaning of the instruction. Thanks again for writing it.
· Reply · 5 d · Edited
Steve Goss
Thanks for this David, great stuff. Of course, the orchestral version was written after Bream had performed and recorded the piece. There are several nods to Bream in the orchestral score — the key you mentioned, some percussion effects in №3 and the slowdown in 1. Bream was almost coercive in his collaborations with composers. Perhaps he convinced Walton that slower was better? Walton keeps his own tempi and dynamics in Bagatelles 2 and 3 in the Varii Capricci. So maybe we can conjecture that the Varii Capricci markings and changes in some way represent Walton’s responses to Bream’s interventions. Late in life, Walton’s confidence as a composer was at Rock bottom, so maybe he was more susceptible to influence?
· Reply · 6 d · Edited
Steve Goss
Then there’s the whole Malcolm Arnold business…
· Reply · 6 d
Raymond Burley
A very interesting article David. We know Bream played the pieces (or at least some of them) to Walton so presumably Bream’s many alterations to the score were approved, although it’s odd that the incorrect rhythm in no.3 wasn’t picked up.
· Reply · 6 d · Edited
David Harvey
Raymond Burley I’m sure the collaboration was as close as it could be in the early 1970s (and we forget just what that was like in the days before email and video chat…) The MSS that’s doing the rounds is a neat copy, not a sketch, and although it differs in detail (and in the key of IV) from the published copy, it’s eminently playable and very close to the published edition, so I’d assume an amount of interaction between composer and performer prior to its creation. I can’t imagine Bream (as I can Segovia) changing things for publication without reviewing with the composer, so the edition reflects an even more advanced state of the collaboration.
· Reply · 6 d · Edited
Raymond Burley
David Harvey
I’ve had that copy for a good few years. Very interesting to see the revisions.
· Reply · 6 d
David Harvey
And this is just joyous: https://www.youtube.com/watch?v=q7eAFBjdYZo
Julian Bream visits William Walton and Susana Walton on Ischia, Italy
· Reply · Remove Preview · 6 d
Ronald Pearl
David — wonderful article! And it shows (again) the value of listening to the composer, and to paying attention to what is written. Raymond brought up the 3rd movement, and the rhythmic alteration — I can remember so many performances of it, à la Bream, that the first time I heard it as written I nearly gasped. But that movement, and the one of which you write, shows the power of a recording: we don’t read any more, but rather find the pitches and mold them to the rhythmic template established by what we have heard (in a way similar to tablature of guitar solos that are published without any rhythmic notation — they rely on the player having heard the recording). It is interesting that the second Bream recorded these pieces, the rhythm in #3 is played as written. (I hesitate to say “correctly” as the unknown factor here is what Walton agreed to after hearing Bream play them.)
Thanks for drawing attention to a detail that is anything but — it seems small, but plays a huge role in the total effect of the piece.
· Reply · 6 d
Mark Delpriora
The passage in question finds an interesting solution… clearly not a mistake::
https://youtu.be/z0A4Qmj59Yw
David Russell: 5 Bagatelles, I. Allegro — William Walton
· Reply · 5 d
David Harvey
Mmmmm I _really_ don’t like the change of gear coming into m.56, nor the sneaky extra note at the end of m.64 … A pity, because up to that point it’s really compelling.
· Reply · 5 d
Mark Houghton
A query to the knowledge present here: should the ‘espress’ in m.66 really apply to m.65 and, by definition, also influence tempo in the section in question?
· Reply · 5 d
David Harvey
Mark Houghton Interesting… In the MSS there’s not a system break at this point. *In the idiom* It would make no sense to play 65 _meccanico_ and 66 _espressivo_ — William Walton is not Elliott Carter…
· Reply · 5 d · Edited
Ronald Pearl
David Harvey I wonder if it isn’t an issue of engraving “convenience.” There is a lot of information in bar 65 — dynamics, position, tempo indication; that’s a lot of vertical space, and the engraver might have thought to add the indication in the next bar, given the texture is the same. Just a guess, though.
· Reply · 5 d
David Harvey
Ronald Pearl Indeed, but in the MSS the _espress_ is also in m.66
· Reply · 5 d
Ronald Pearl
David Harvey Perhaps as an intended increase in intensity/expressivity; the change in texture and rhythm is already an expressive gesture — adding the espressivo is a way of saying “keep it up, don’t just repeat the idea the same way.” But then who knows — it’s in Italian. if he really wanted us to know what he wanted, he’d have written it in English😉.
· Reply · 5 d
David Harvey
Your point is valid for the MSS as well as the printed copy, but the system break makes the ‘espress’ appear a bit more separate
· Reply · 5 d
Ronald Pearl
David Harvey Too bad they didn’t have modern engraving software back then…😉
· Reply · 5 d
Mark Delpriora
You know, after listening to a bunch of performances, I would say guitarists have rejected the composer’s marking. It’s a rebellion!!
· Reply · 3 d
Ronald Pearl
Mark Delpriora Composers. What do they know? They act like they wrote the piece.
· Reply · 3 d
Mark Delpriora
Few follow the metronome marking in number 3!, too!
· Reply · 3 d
David Harvey
A shade slower in the orchestral recording! Changes the character completely.
· Reply · 3 d
David Harvey
https://youtu.be/tT1cTEsx0_0?t=400
William Walton (1902–1983) : “Varii Capricci” for orchestra (1976)
YOUTUBE.COM
William Walton (1902–1983) : “Varii Capricci” for orchestra (1976)
William Walton (1902–1983) : “Varii Capricci” for orchestra (1976)
· Reply · Remove Preview · 3 d
Mark Delpriora
Yes! I like the notated tempo.
A real issue is that students are afraid to follow the score if it is not common practice. I had a student play # 3 for a well known personage, usually a follow-the-text guy, and he skewered him for followed the marking…… See more
· Reply · 3 d
David Harvey
There are a couple of recordings in the vicinity of the notated tempo (props Sean Shibe!) It does (as Jonathan Leathwood pointed out to me) change the character of the whole set — the three inner pieces gain a nocturnal (sic) intensity that’s completely dissipated if III is interpreted as a nervous scherzo…
· Reply · 3 d · Edited
Mark Delpriora
That makes sense about the inner 3 pieces.
· Reply · 3 d
Ronald Pearl
Mark, David- There is a story that Gary Graffman tells in his autobiography about playing Schumann (?) for a jury, and ending the piece with the dynamics as written, and not as was — at the time — performance tradition. One juror asked another, “is that how it’s actually written?” The reply from another juror was, “yes.” They still hammered him…Some things get a life of their own — like the extra measure that was added to the 1st Prelude of the WTC. I have an edition of it by Fauré, and over the measure in question is an asterisk; the asterisk states (loose translation),” not in the original, but consecrated by tradition.” So, if we’re willing to do this with Bach, then really no one is sacred.
