Table 2.5 summarizes the row-level activity of the first movement. The exposition moves from the ‘tonic’ row level to row level 4, the development moves through various row levels, and the return moves from the tonic row level to level 4 and finishes on the tonic row level.
After the rather detailed account of the first movement, the following two movements will be analyzed more summarily.
One of the ways that Dallapiccola differentiates the dream state of the middle movement from the real action in the rest of the piece is to employ PB, a row unrelated to the rows of the of the outer movements. (Note that the row is divided between two octatonic collections that share op5, a procedure that is fairly common in Dallapiccola’s row construction in the 40’s and early 50’s.)
Ex. 2.16 Rencesvals II: PB0
The movement’s arch form, A B C B A, is perhaps inspired by the reference to Charlemagne sleeping that appears at the beginning and end of the stanza. The sense
Ex. 2.17 Rencesvals II, sections A and B: row levels
of drifting off to sleep is evoked by the long chord that is held over from the first movement into the second as well as the rhythmic decelerando in section A. In mm.46-49, the repeated low D’s become progressively longer, increasing in value by one eighth with each new attack while the notes of the held chord, Bb-C#-D-G are released one by one. (See Ex. 2.17 on the previous page.)
Ex. 2.18 Rencesvals II, section C: row levels Each of the first four sections, A, B, C and B, is distinguished by a different twelve-tone texture and restricted to either prime, inverted, or retrograde and retrograde inverted forms of the row. In Section A, a linear presentation of one row form, PB0, is split between voice and piano. Section B is a canon between two row forms, IBt and IB5. In section C, the soprano and piano split R6 and RI4, which follows R6 with some overlap. Section B is a transposition of section B. (Since this section is textless, one concludes that, in this instance, the musical design is independent of the poem- unless one interprets section B as Charlemagne’s unstated desire to reconstitute the spear.)7 Finally, section A is a transposition of section A with elements of sections B and C added. The last section presents various transpositions of PB, IB and RB, at first concurrently, and then consecutively. The row presentations become more chordal, with a return to linearity in the last three measures. Section A, therefore, constitutes the textural culmination of the movement.
The shattering of the spear in section C is expressed by the textural discontinuity of the piano accompaniment. The explosive action of these verses is evoked by the increasing agitation of the tempo (quarter note = 4072) while the dynamic level remains pianissimo throughout, suggesting that the action is not real.
Although the movement briefly hints at an ending on the ‘tonic’ row level with RB0 in m.66, there is no tonal analogue; the material in sections A and B returns in sections B and A at different transpositional levels. The absence of the original row levels at the end of the arch form projects an unbalanced symmetry that expresses the fluid nature of the dream.
Table. 2.6 Rencesvals II: row levels
55 57 58-60
63 64 66 67-70
IB2 PB4 RB0PB8
Because D is the held bass tone of section A (see mm.46-51), by ending the movement with PB8 rather than PB0, Dallapiccola is able to finish with the held bass tone Bb. As Ex. 2.19 demonstrates, this permits section A (see mm.63-69 in the score) to be heard as a linear echo of the bass Bb at the end of the first movement.
Ex. 2.19 linear graph of Rencesvals II in relation to Rencesvals I Rencesvals III
The piano accompaniment in the third movement returns to the texture of three-chord modules typical of the first movement. Example 2.20 demonstrates the close relationship between the modules MA and MB of the third movement and M of the first movement. If C and C# are exchanged MA0 is transformed into MB0 and vice versa. If chords two and three are reversed and the E and Eb exchanged, MB0 is transformed into M5 and vice versa. MA and MB are in the closest relationship, followed by MB and M. MA and M (two pc exchanges and a chord reversal) are the furthest removed.
Ex. 2.20 Rencesvals: modules
Dallapiccola achieves a motivic connection to the first movement row, PA, by spacing the modules MA0 and MB0 so that the bass notes form the motive B-G-A
(-4 semitones + 2 semitones). (See Ex. 2.21 on the next page, m.70.) This motive is an intervallic expansion of the chromatic neighbor note motif Eb-C#-D that appears in m.1 as the first three pcs of PA0.
Even though Dallapiccola uses similar modular material in the first and third movements, he imbues them with characteristics appropriate to their respective stanzas, by means of rhythm and register. The deep valleys of the third movement are suggested by the registral chasms that open up between MB0 and RMB1 in mm.75-80 (see Ex.2.21), and MB1 and RMB2 in mm.88-93. To convey the solidity of the mountains, each chord takes on the value of a whole note. There is very little that
Ex. 2.21 Rencesvals III, mm.70-84: row levels could be construed as a melodic line in the accompaniment; it is monolithically chordal. Nor is there an independent linear presentation of row material in the voice. Its jagged line, outlining peaks and valleys, is extracted from the piano harmonies.
As demonstrated in Ex. 2.20, MA and MB are similar enough so that their row levels can be considered equivalent. Table 2.7 illustrates the row level structure of the third movement.
Table 2.7 Rencesvals III: row levels
pno. & sopr.
MA0 MA0 RMA0 RMAe
MB0 RMB1 MB0
pno. & sopr.
MA4 RMA3 MA3 MA1 RMAt
MB1 RMB2 MB1
MA5 RMA4 MA4 MA2
MA1 RMA2 MA2 MA0
Section A establishes a ‘tonic’ row level. MA0 first appears in incomplete forms, as the bass line of the three-chord progression only (mm.70-71), and then as the bass and tenor voices only (m.73); the first complete module at the ‘tonic’ level is RMA0. In mm.75-84, the ‘tonic’ level, MB0, brackets MB1. (The major ninth, B-C#, in m.70 is held over from the second movement similar to the way the chord Bb-C#-D-G is held over from the first movement into the second.) However, as in the second movement, the return to the tonic level (m.103) is only a feint before the conclusion on the ‘e’ level, MAe. The long-range motion MA0 MAe is prefigured in the opening piano solo, MA0 RMAe (mm.70-74). This long-range motion is recapitulated at the end of the coda, MA0 MAe (mm.104-105). The global linear motion of the entire piece, Bb-G#, is also recapitulated between measures 103 and 105. See Ex. 2.22 and the linear graph that follows on p.60, Ex. 2.23.
Ex. 2.22 Rencesvals III, mm. 103-110: row levels Example 2.23 shows that the A-G# first appears in the opening piano solo (mm.70-74) where the G# is heard as auxiliary to the A. However, the two replications of the structural A-G# motion, and the final Bb-Ab that follow (mm.77-82, 95-100 and 103-05) make it clear that G# (Ab) is the goal tone. Example 2.24 demonstrates the global motion of the entire piece. Thus, although it very difficult to trace a tonal analogue through all three movements owing to the different rows used in each movement, there is a linear thread running throughout the piece. Like a great sigh, the whole-tone long-range descent expresses the poem’s sense of impending doom.
Click here to view Ex. 2.23 and Ex. 2.24
1 Owen, The Song, pp.13-14.
reconstruction suggested by Dallapiccola’s friend Prof. Luigi Foscolo Benedetto
2 Nathan, “On Dallapiccola’s”, p.35.
3 Nathan, “On Dallapiccola’s”, p.35.
4 The movement ends with a three-measure extension of the chord played by the left hand in m.42, further emphasizing the return of that harmony which first appeared in mm.1-2.
5 Berlioz does something similar in the third movement of the Symphonie Fantastique. The second theme is in the dominant, C. In the recap it comes back in C over a C pedal before the movement ends in F.
6 The attempt to reconcile the harmonic and linear structure of Rencesvals I by identifying the final chord, Bb-C#-D-G, with a G root is ill-advised. A chromatic chord is always ambiguous, and in this case, a vii chord in B minor would be equally, if not more, plausible!