Alec Rowley’s “Lied,” Op. 42, No. 13
A Performance Analysis
“Lied” is a character piece from Rowley’s collection, 30 Melodic and Rhythmic Studies. This analysis aims to give advice as to how to interpret it.
This piece has an introduction that concludes at the end of m6, and a coda that begins in m39.
The conclusion of the introduction is marked with a ritardando (m6). The conclusion of the body of the piece (m38) isn’t, although I’d suggest inserting a point of relaxation there as a matter of interpretation. In a sense, Rowley has encouraged this: he avoids an articulation of the third beats of m37-38. At the same time, he presents two back-to-back dotted half notes in the melody with no (or at the very most, negligible) secondary melodic activity in the accompaniment. In essence, he’s writing out a subtle ritardando, or something closely-related to one.
I hear the introduction as a 4+2. However, the crescendo in m3-4 – because it leads all way to the downbeat of m5 – compels me to think of the entire intro as a single unit. This crescendo is potentially problematic in that it seems to conflict with the resolution of the appogiatura in the alto voice, which – at least to my way of thinking – calls for a diminuendo. I can get my mind around that by considering the crescendo to apply to what’s happening globally, while allowing localized dynamics to occur independently. However, in a parallel passage (m39-44), this crescendo is only half as long, and much more in line with my instincts. This leaves me with questions:
- Is this an error, and the crescendos should be the same?
- Is this deliberate, and I should strive to differentiate these passages?
- Am I really going to spend this much time and energy obsessing over a small, Alec Rowley piece?
The answers are yes, yes, and yes. The solution I’ve come up with is a compromise that allows for those yesses. I’m going to use a single-measure dynamic crescendo in both passages, but in the first of them, I’m going use pacing that pushes all the way through the passage, and in the second, I’m going to relax at the close of the crescendo.
In further support of differentiating these passages, I’d like to point out that m6 and m44 are less alike than you might think at first glance. The ritardando, sure … but the soprano – which splits into two voices in the earlier measure, but doesn’t in the later one – offers even more compelling evidence of these measures’ differences. In m6, the split soprano creates a textural crescendo that parallels the dynamic one. Furthermore, you cannot play the split note with a legato touch, even though it’s under a slur, because of the repeated note that precedes it. Instead, you are – in all likelihood – going to lift your hand at the end of the first beat, and land on the second. In every way that he can, Rowley is asking you to emphasize b2 in this measure instead of b1.
The way that m6 and m44 differ most, however, is that m44 is not the terminal point within its unit. Emphasizing b2 forces you to think on the quarter-note level, which slows the pacing. This works perfectly in the introduction. However, the key to the coda is to carry it through to the final three measures, which requires you to think in whole-measure “beats” until you get there. You have to observe the dynamic markings in m44 while still emphasizing the downbeat … perhaps by relegating them to the accompaniment.
The phrases within the body of the piece work in a hierarchical way:
The slurs indicate one-measure units. It seems logical to group the slurs into twos: m6-7 make an insightful unit, for instance. Likewise, the two-measure units seem to group into four-measure units, and the four-measure units also form groups of two, which leaves us with eight-measure groupings – and I think that’s where we stop.
All of these groupings have a bearing on how you play the piece. Subtly emphasize the first note of each measure (one-measure units). Play the grace notes in a way that draws more attention to the measures that contain them, and deliberately diminuendo to the end of those measures. (This creates two-measure groups). Dynamically elevate m9-10 only as much as necessary to differentiate them from the previous two measures, while at the same time maintaining the same dynamic plan you established in those previous measures. Do the same thing in m11-12, this time with an obvious crescendo; however, instead of a diminuendo in m12, allow the crescendo to crest on the downbeat of m13. This turns m11-14 into a four-measure idea, which retrospectively turns m7-10 either into another four-measure group, or a two-plus-two.
One good reason for thinking of m7-14 as an eight-measure idea is the lack of a distinguishing cadence in m10 – otherwise, you might want to think of this section as a four-plus-four. Another reason that supports an eight-measure grouping is that long phrases, and the sense of sweep that goes with them, create more of a Romantic feel than shorter groupings would. However, “two-plus-two plus four” has a lot going for it, too. It seems more in-line with this section’s sequential structure. In addition to that, it adheres to the short-short-long Romantic pacing template.
Rhetorically speaking, the body of the piece strongly indicates a climax and conclusion at the end of m30.
The ritardando that appears in m29 is meant to assist with that. However, instead of getting a strongly-anticipated tonic on the downbeat of m31, we get something akin to a deceptive cadence. Rowley uses this device, as many composers do, to retreat, as it were, and take a second pass toward a conclusion, this time delivering on that expectation exactly where you’d expect it to be, in keeping with eight-measure units (m38).
I’m not convinced by Rowley’s use of slurs.
That’s not quite true; the truth is a bit more complicated. What I’m not convinced of are the places that don’t have slurs. For instance, the absence of a slur in the m4’s alto line seems like an oversight. This is where you have to be careful: although a slur always means to play legato, the absence of one doesn’t always mean the opposite. Notice that this piece contains no slurs whatsoever in the alto. There are two reasons for this: first, the slurs in the soprano can be taken to apply to the entire texture, or at least that of the treble clef. Secondly, many composers omit articulations in the accompaniment in ordinary circumstances. Articulations would tend to elevate its importance and make the music sound more contrapuntal than intended.
Rowley treats the bassline the same way: he articulates it, but not the tenor. This is a little funky. For instance, how could you not hear m3-4 in the bass clef as an extension of the previous tenor melody? Clearly, it’s the bass that drops out, not the tenor. I agree with the slur, but it really ought to begin in the first measure, over the tenor. However, doing so would invite the player to wonder why the rest of the tenor line is devoid of articulation, so I get it. (But it’s misleading. It’s not easy for a composer to balance these kinds of ideas.)
Pitfalls
- The quarter note is not the beat! Playing it that way will expose you as a student, rather than an artist. The downbeat is the only beat; you must treat beats two and three as offbeats.
- The melody is the Beyoncé of this piece. The other notes are her backup singers. If you make me have to expend any effort whatsoever to distinguish her from her backups, I will come down so hard on you with a Twitterstorm of immeasurable magnitude that you will not survive.