David Victor Feldman: Viola Sonata

First contact and getting the score

My introduction to David Feldman’s music was through listening to a MIDI playback of his "A Study in Contour (1975)" which is the fourth movement of his Sonata. I rarely bother listening to MIDI simulations, but there was something that caught my attention in the melodic construction. I later realised that the first eight notes of the Study in Contour were very similar in contour to a melody (at bar 45) in the third movement of my Serenade for Strings (2001). The pitch was close, too - quite a classical configuration. After getting the score (already computer-set) I started working on this Study sometime early in 2018, and during the second lockdown of 2020 brought it out again to make a video of this movement. I was then sent a copy of the manuscript score to the first three movements of the Sonata which I set in Dorico before starting working on it.

The composer’s background

It’s been interesting discussing aspects of the Sonata over the last few months and the following paragraphs are a selection of some of the points raised. Knowing that David must've written the Sonata at age 17 or 18, I was curious to know more about his musical education, and what instrument he played (or had played).

DVF: I starting playing the flute at about age 8. At about age 14 I decided that I wanted composition lessons. A couple of years before that, a school teacher (with whom I am still in touch!) had exposed me to Pierrot Lunaire, and to Webern, and that was the real start. A composer, Leo Kraft, from the town where I lived, took me on as a student. After a couple of years of lessons with Leo I went off to college at Yale. The philosophy at Yale was that each student should work with as many different composers as possible. Since I graduated in three years, I had six different teachers. Penderecki is the name everyone knows, but those lessons came a year after I composed this piece. Indeed, I composed this piece while studying with Robert Morris. Bob then was interested in polystylism and acculturation, but he was also already a great expert on Babbitt-flavored serial technique.

I wasn't composing 12-tone-music, but this piece is based on a row of 24 overlapping 4-note contours, an idea I'd invented myself. (While one can make a 12-tone row by fiat, but it's not even mathematically obvious a priori that these contour-rows exist.) One can recognize a contour even with no ear training in relative pitch, so the idea was a move towards a more accessible serialism.

The idea, very novel at the time I think, was to serialize...not pitches...but contours...There are 24 possibilities with 4 distinct notes, and I created a cycle that travels through all 24 in 24 notes. (For example, one contour would be three rising intervals, another three failing, etc.) The democracy between all the contours creates a kind of melodic vagueness, since contour often unites variations, non-literal repetitions. I don't know how this affects the performer except...let yourself react to the rising and falling intervals, perhaps with dynamics or rubato.

The manuscript didn’t have time signatures, and I’d put them in, somewhat because I was still relatively new to using Dorico and still feeling my way with this software.

DVF: Most players do prefer time signatures. I never used to put them in because they have common practice connotations concerning phrasing, for example 6 8 versus 3 4, and often my rhythm concept is additive rather than divisive. But that was a long time ago, my music has changed and players have changed, and I think people get it now when I say "a time signature is just a number (or two)." Whatever works.

The overlapping triplets is one example of the additive approach. For example, a new triplet starts on the third note of the preceding triplet. To achieve this in Dorico, rather than have triplets overlapping, it was set using the 5/24 metre. The occasional high C-sharps after previous A-sharps in a similar passage prompted me to verify the pitches.

DVF: Well three things: the manuscript is clear; as a flutist ledger lines wouldn't tend to confuse me now or then; I had definitely picked up the idea of "long-range atonal voice leading" and specifically from late Stravinsky how effective it can be to occasionally have the movement by a third rather than a step. So I think it's right/it should stay.

Stravinsky is a composer who I’ve admired for a long time. There’s also something in the overall structure and writing of the Sonata that makes me think of the Stravinsky Violin Concerto.

There’s a recurring tremolo between two double-stops (bar 55) that at first I thought was going to be problematic;

NK: At bar 55 what's the intention there? It's not possible to do a fast double stop between the two sets of double stops as the hand position is too awkward, in part because of the D being an open string.

My initial reaction now seems quite exaggerated! It was simply a question of putting fingers 1 & 3 on the lowest two strings, then remembering to pull the 2nd finger on the C# to the outside of the fingerboard to allow the interior open string to vibrate.

DVF I just imagined, essentially, a triple stop + open string. So the effect is mostly accomplished by bowing. Not quite an arpeggio. The bow would rolls from playing the two low strings to the two high strings. In between of course it would catch the two middle strings. I'm not imagining a clean effect, so nothing like what the notation would mean on the piano. The fingering doesn't require any big stretch, right? Fingers 1,2,3 or perhaps 1,2,4 could reach it.

The pencil score has "crumbandos.” It's a way of indicating an accelerando that George Crumb either invented or made popular for a time, but I think they've fallen out of fashion

Matters of interpretation

Aside from specific technical matters needing solutions, it’s been a challenging piece to interpret, because of the diverse rhythmic ideas in close proximity. Sometimes driving (even rock-music influenced) rhythms shift quickly to quite temporally-suspended passages. It’s a very physical work to play, very colourful and inventive with lots of extremes.

NK: My vision of the first movement interpretation is quite clear now. I see it as having a base tempo throughout. It follows this initial tempo until bar 41 inclusive. 42-54 is a qusi-cadenza. 42-44 is the first thematic phrase of that section. It comes back to the tempo primo at various points from 55 onwards. Instinct in interpretation is somewhat paradoxical, I find, the more organised it gets the freer it feels.

I'm searching most for ideas on the rhythmic continuity. Do you have any comments there?

DVF: If I were going to make a suggestion, I would say...feel free to make it more impulsive, instinctual. Don't worry about "dragging" and "rushing." The pitches are all about "voice leading" and if your ear hears that a note connects to some note in a measure or two previous, you can let that inform the continuity (through matching dynamics or articulation...or I don't know what). I would say think "fragmented, interrupted dance." So that kind of forward impulse, but the music interrupts itself, hesitates, rethinks. I often say to players that they should think of my score as a still-life set-up for painter, not for a photographer. Truly I'd rather have you following your instincts that feeling enslaved to markings.

A thought that has long occupied me, but admittedly not as long ago as I wrote that movement, but perhaps still relevant: good actors have no inertia. They respond immediately when they need to...so an emotional body in motion doesn't tend to remain in motion...unless that serves the drama. To the extent that musical interpretation can be like a kind of acting, one can get an extraordinary effect with modest but *immediate* pivots...so that the feeling of one musical thought doesn't bleed into the next just because of proximity in time. This is a challenge for some performers because of course one does physically move from one thing to the next, so some sort of transition is natural. It's a hard exercise, say, to play every phrase of a piece like it's the first phrase.

NK: I'm thinking the opening of the sonata should be marked at a much lesser dynamic, I'm starting to see it as quite lyrical, to reserve the gritty stuff for later.

DVF: Let me add...I'm 64...the standards of new music performance has risen dramatically since I first started composing. As a student I would hear concerts in NYC and the performers just seemed very tight and uncomfortable. They weren't particularly trained to play new music, so it was a "second language," and then the abundance of markings deprived them of all agency, so the performances were about as personal as watching an office typing pool. I guess I'm still reacting against that, even though there's a generation of young players who do now play new music with gusto.

Two things: First, I have learned that a MIDI interpretation is a thing unto itself...one has to rethink a piece because, for example, a MIDI viola is just not a real viola. Second, I rebel against the, shall we say, post-Darmstadt tradition of score that project very precise intentions. I really love it when interpreters *interpret*. Please play my pitches and rhythms, but I'm open to very liberal interpretation of dynamics and even articulation. The score shouldn't be a prison...feel free to make the best experience for yourself and your audience. I often say I'm open to having my pieces "covered" (like they say in the context of pop) as much as "performed." Please treat my markings as serious suggestions, but especially since you're an excellent composer yourself, don't deny those instincts!

The pitches are all about "voice leading" and if your ear hears that a note connects to some note in a measure or two previous, you can let that inform the continuity (through matching dynamics or articulation...or I don't know what).

I just wished I'd had a player to work with when I wrote it. At this point, having heard what you can do, both as player AND composer, I totally trust your instincts.

Feel free to edit the score so it conforms. So when I put the score out in the world it will match your performance. And solve problems if anyone else ever takes it up.

I can't explain what to do. But I can say something vague. The listener hears events...the notes in this case. In reality you cause all the events by playing the notes. In the fiction of the music, perhaps some notes seem to cause or at least affect one another. When the rhythm gets performed perfectly evenly, it sounds a little frozen. Of course it warms up as it develops anyway. So it works for me either way. But some micro-asymmetry would have it start just a little warmer. One solution might be to take a note or three to get to tempo. Just a little instability to draw in the listener. If I could explain it I could have notated it, but the variation I'm imagining is so subtle that it would be unnatural to notate.