Concerto for Two Double Reeds in detail: IV — Rondo

This is the fourth in a series of post-performance analyses of my Concerto for Two Double Reeds. You can also read about I — TessellationsII — Meta-Canon and III — Loops. Click here to read the full score of movement IV — Rondo or listen to the live recording.

The fourth movement is an old-fashioned rondo with lots of tunes; however, the original idea was rhythmical rather than melodic, and these rhythmic concerns recur throughout the movement.

Perhaps the decisive moment for me in writing this movement was when I noticed a rhythmic parallel between a Latin poem and an Italian canzonetta.

Catullus 63 treats of a subject that wouldn’t be out of place in an Almodóvar film: a young man, Attis, a devotee of the cult of Cybele, enters a trance and emasculates himself:

Super alta vectus Attis celeri rate maria
Phrygium ut nemus citato cupide pede tetigit
adiitque opaca silvis redimita loca deae,
stimulatus ibi furenti rabie, vagus animis
devolvit ili acuto sibi pondera silice.

Aside from its rather provocative theme, what marks this poem out is its driving rhythm:


This rhythm is known as the Galliambic, and the first half (before the caesura — marked ‘||’) is also known as the Anacreontic. The Anacreontic rhythm can be considered as the result of anaclasis (inversion) of the middle two syllables of an Ionic dimeter:

⏑⏑––|⏑⏑–– becomes ⏑⏑–⏑|–⏑––

Now, if you go back and read my old teaser post for this concerto, you’ll see I discuss the patterns of bell ringing, which are based on swapping values in a series. The transformation from the Ionic dimeter to the Anacreontic follows the same technique (albeit in a fairly limited fashion).

The second source is the canzonetta Vi ricorda o boschi ombrosi  from Monteverdi’s L’Orfeo:

Musical example illustrating the Anacreontic rhythm in 'Vi ricorda o boschi ombrosi' from Monteverdi's L'Orfeo

Here is the same passage in modern notation:

The same excerpt in modern notation

The pervasive rhythm of this passage is again the Anacreontic:


I had been keen for a while to play with classical rhythms, and this observation prompted me to start playing with the technique of anaclasis.

The opening passage of this movement doesn’t apply this technique to the Ionic dimiter, but rather to an Anapaestic rhythm:


After anaclasis, this becomes the following rhythm:


Join the two rhythms, and you get this:


This is the rhythm of the opening phrase of the Final of my concerto:

Opening oboe theme, illustrating the anaclastic anapaestic rhythm

In addition to introducing an important rhythmic technique, this opening passage also sets out two other important aspects of the movement:

  1. The anaclastic rhythm cuts across the barline (the A♯ quavers tied from b.3 to b.4) giving a syncopated accent;
  2. The melodic material is derived from triads on C and F♯, which lie a tritone apart.

The entire first phrase shows how these three ideas are developed:

The entire first phrase. The two-bar syncopated rhythm is repeated slightly lower, and then finished with a falling anapaestic rhythm

This brief opening fanfare is played in the oboe and bassoon, and leads into the rondo theme at [A]:

The rondo theme: two phrases is straight/anaclastic anapaestic dimeter, then a phrase in straight/anaclastic ionic dimeter

The first eight bars of this theme are based around the Anapaestic dimeter rhythm already explained, but with a small variation in the second foot. The following four bars, in 3/4, oppose a straight Ionic dimeter with its Anacreontic variant. This material maintains the polytonal nature of the opening: the strings play a vamp rhythm alternative between C and F♯ chords every two beats — even after the change to 3/4 —, and the thematic material meanders between these two tonal centres:

The rondo theme with accompaniment

This material is interrupted with a harsh outburst in the strings, again varying the Anacreontic rhythm across an interval of a major 7th, which is then picked up by the oboe in the inversion of that interval: an minor 2nd, leading back into the rondo theme, which is rounded off with a descending variant of the broken chords that opened the movement.

Episode I at [E] takes the ideas of the rondo theme and explores them in a different context. The alternating chords are each moved by a semitone, from C and F♯ to C♯ and F. The vamp bass remains, but the upper strings now play counterpoint above it, rather than pizzicato, and the solo material is a duet between oboe and bassoon. The technique of anaclasis appears again, this time transforming a dactylic dimeter (which is not explicitly heard):

–⏑⏑|–⏑⏑ becomes –⏑–⏑⏑⏑

Concerto for Two Double Reeds IV: Rondo, first episode

This episode leads straight back into the rondo theme at [F], which is this time finished with arpeggi in the low strings, which recall the solo material at [E].

Episode II leaves behind anaclastic meters, and explores Aeolic rhythms instead.

The core rhythm of Aeolic meter is the choriamb:


This basic unit is extended by the addition of syllables before and after. The syllables before are anceps (either long or short), while those after tend to alter breve, longum. For instance, the rhythmic colon known as the Hipponactean looks like this:

⏓⏓ –⏑⏑– ⏑–×

In this episode, I decided to work freely with the principles of Aeolic rhythm, rather than adopting a preexisting verse form. Here is the resulting rhythm:

–– –⏑⏑– ⏑–⏑–
–– –⏑⏑– ⏑–⏑–
–– –⏑⏑– ⏑–
⏑⏑ –⏑⏑– ⏑––

In each of these lines the central Choriamb can be seen clearly.

The Episode II theme in the bassoon (and violin) at [G]

This is the first episode in which the bassoon is the only soloist, and the first to move away from tonality. Rather than diatonic scales, the melodic and harmonic material is drawn from a scale that Messiaen would have described as the 2nd Mode of Limited Transposition, and which is also known as the octatonic. This is made of alternating tones and semitones.

Octatonic scale on C

The bassoon explores this scale, while the upper strings play diminished triads — the chord that forms if you take alternate notes from this scale. As in the rondo theme, there is an ambiguity between 2/4 and 3/4 meter, but this time it is the string figurations that change:

The rhythm shows two 3/4 bars, one 2/4 bar; two 3/4 bars, one 2/4 bar; two 3/4 bars with a hemiola; two 2/4 bars; one 3/4 bar

The material of this episode is played three times:

  1. Bassoon solo with upper string triad accompaniment;
  2. Oboe solo, bassoon counterpoint in diminished arpeggi, upper string accompaniment with cello bassline;
  3. Oboe and bassoon have the melody, the upper string triads are spaced by octaves, the double bass joins the bassline.

The material then dissolves until only stratospheric violin and abyssal double bass are left. At this point a solo violinist picks up with the material first heard in the interruption to the initial rondo theme, which then leads into a statement of this theme in solo violin and pizzicato viola at [J]:

Rondo theme in solo violin and viola

With the soloist played at such a high pitch, and no harmonic accompaniment, this material sounds very different from its initial statement. The material is then picked up in both violin parts (tutti), with accompaniment from all the lower strings, but still no harmony, and we then then move into the next episode at [L].

The rhythm of Episode III is again Aeolian. A feature of Aeolian rhythms is that the central Choriamb can be expanded, either by full repetition or with Anapaests (which are equivalent to a Choriamb with the first syllable missing). This passage uses Choriambic expansion, moving between Glyconics:

–⏑ –⏑⏑– ⏑–

And their expanded equivalent, the Asclepiad:

–⏑ –⏑⏑– –⏑⏑– ⏑–

The combination Glyconic, Asclepiad, Glyconic, Asclepiad is known as the Fourth Asclepiad, and is used by Horace in his Ode 3.9:

–⏓ –⏑⏑– ⏑×
–⏓ –⏑⏑– –⏑⏑– ⏑×
–⏓ –⏑⏑– ⏑×
–⏓ –⏑⏑– –⏑⏑– ⏑×

Donec gratus eram tibi
nec quisquam potior bracchia candidae
ceruici iuuenis dabat,
Persarum uigui rege beatior.

The rhythm I use for this section is slightly different, being composed of two Glyconics followed by an Asclepiad:

–⏑ –⏑⏑– ⏑–
–⏑ –⏑⏑– ⏑–
–⏑ –⏑⏑– –⏑⏑– ⏑–

Also, unlike Horace’s practice, I chose to use a short value on the second value of each line, and to keep the final value long; this keeps each line absolutely symmetrical, which adds to the character of this episode.

Oboe theme of Episode III

The melodic material for this episode is based entirely on whole-tone scales, alternating each bar between the two possible transpositions of this scale. The solo material makes a lot of the major 3rd that emerges from this scale, while the accompaniment, spaced in major 3rds, traces consecutive notes.

Accompanying figuration in Violin I

This episode falls into four sections:

  1. Melody in the oboe, scales in violins;
  2. Melody in the bassoon, scales in viola and cello;
  3. Melody in violin II and viola, harmonics in cello and double bass, scales in oboe and bassoon;
  4. Scales in oboe, bassoon and upper strings, harmonics in cello and double bass, then inverting roles with the upper strings playing very high notes and the cello and bass playing scales.

After these discursions into non-diatonic territory, we return at [N] to the final statements of the rondo theme. This time it is stated in the bassoon,  and is fully harmonised.The interruption from the first statement reappears here, with the oboe’s lead-in slightly altered, and then the theme reappears in counterpoint at [P], first in the bassoon with bar-spaced oboe arpeggi, then at b.297 in the oboe with constant bassoon arpeggi. These arpeggi take over at b.305 in both soloists and upper string, leading into the final statement of the theme at [Q], in both oboe and bassoon, with the string pizzicati replaced with arpeggi:

Rondo theme plus C/F# arpeggi

At b. 321 the soloists abandon the rondo theme, picking up the movement’s opening material, and are joined by the entire orchestra playing this fanfare at b.329, before the movement closes with exact material that opened it, but across the whole ensemble:

Opening theme brought back tutti as the coda.

I have to confess to feeling slightly ambivalent about this movement. It is unquestionably a crowd-pleaser, and I have a feeling it’s good fun to play, but its very richness in tunes and compelling rhythms raise nagging feelings that it’s rather superficial and not serious music. Perhaps the analysis on this page is something of an apologia for it: an attempt to show that there are actually plenty of clever ideas behind its appealing tunefulness and vampish character. Or perhaps I should just relax and enjoy it!

(*Please read my note on copyright.)

Concerto for Two Double Reeds in detail: I — Tessellations

This is the first in a series of post-performance analyses of my Concerto for Two Double Reeds. Click here to have a look at the score of movement I — Tessallations or listen to the live recording*.

The first movement of my Concerto for Two Double Reeds is for oboe solo and orchestra; it builds on two ideas:

The first idea is stolen from György Ligeti, and it is to divide the chromatic scale according to the keys on the piano: the white keys form a diatonic scale, and the black keys form a pentatonic one. In this case, the white notes are given to the oboe, and the black notes to the orchestra.

The orchestra play their notes with two textures: sustained chords, and descending pizzicati passed from instrument to instrument. The five notes of the pentatonic scale are divided between the five parts: Violin I, Violin II, Viola, Cello, Double Bass in a succession of narrowing, then broadening spacings:

The following chords:  [͵A♯ G♯ f♯ d♯´ c♯´´] –  [F♯ c♯ g♯ d♯´ a♯´] –  [c♯ f♯ a♯ d♯´ g♯´] –  [g♯ a♯ c♯´ d♯´ f♯´] –  [c♯ f♯ a♯ d♯´ g♯´] –  [D♯ c♯ g♯ d♯´ a♯´] –  [͵A♯ G♯ f♯ d♯´ c♯´´] –  [͵D♯ D♯ d♯ d♯´ d♯´´]

Throughout this sequence the second violins hold middle d´♯, which acts as a pivot for the other notes. Within the pentatonic scale, the first chord can be considered to be build on 5ths (A♯ – C♯ – D♯ – F♯ – G♯ is five successive degrees of the pentatonic scale), the second on 4rds, the third on 3rds, and the fourth on 2nds, so the spacing is entirely regular, albeit within an irregularly spaced scale.

The second idea is to base the oboe part, within the confines of the diatonic scale, around a repeating group of intervals. All the oboe material is based around a ladder, or tessellation, formed of rising perfect fifths and falling major seconds, a technique which acknowledges Messiaen’s modes of limited transposition, whilst doing pretty much the opposite!

Treble clef, in succession the notes e´ – b´ – a´ – e´´ – d´´ – a´´ – g´´ – d´´´ – c´´´

Pushed much further this pattern would encounter black notes, but then again, pushed much further it would exceed the oboe’s range!

The resulting movement is in two sections:

In Section 1, the strings hold each, successively narrower, spacing of the pentatonic pitch set while the oboe plays increasingly florid, improvisatory material. For each chord the oboe plays three phrases of increasing length, followed by the pizzicato string motif.

The final, close-spaced pizzicato motif of Section 1 becomes the first motif in Section 2, which is for strings alone. The abandon the sustained chords, and just play the pizz motifs in reverse order. Finally, reach a 4-octave span of D♯s, which they play three times, ending fff to end the movement.

In workshopping this movement, we discussed various ways to notate the oboe passages. I had originally been fairly specific about which notes should be triplets, and about dividing the material into bars, but felt that this detracted from the rhythmic freedom the performer should have in interpreting the material. Rhuti’s performance brought a lyricism to this movement that went beyond what I had envisioned, and even elicited comparisons — unexpected, but not unwelcome — to  Delius.

Note: if you would like to listen to a recording of this performance, please get in touch and I can send you a link.

(*Please read my note on copyright.)

Oboe concerto teaser: II

This is part 2 of a series on my forthcoming oboe concerto. This is a commission for the Handel Collection, and will be performed on 5 July 2011, 13:0014:00 at St Stephen Walbrook (warning: site plays music), 39 Walbrook, London, EC4N 8BN. Please come along if you can!

I mentioned last time that the Sarabande is built on a meta-canon. Here’s how it works.

This movement is 156 bars long which is equal to 12 × 13 or 2 × (1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 + 10 + 11 + 12). This second formula is salient, because it’s based on a triangular number, and I’m rather fond of triangular numbers.

The oboe and bassoon parts are each divided into 12 sections, each of a different length 2n where n is a number from 1 to 12. The oboe’s sections are ordered n = 12, 9, 6, 3, 2, 5, 8, 11, 10, 7, 4, 1, and those of the bassoon n = 2, 4, 6, 8, 10, 12, 11, 9, 7, 5, 3, 1, so they both similar types of pattern.

As the sections are of different lengths, the beginnings of sections rarely coincide in the two parts, and only once on the same length of section: n = 1 in the final 2 bars. Other coincidences of note are that oboe 2 and bassoon 12 start together (mirroring the beginning of the movement, where oboe 12 starts with bassoon 2), bassoon 11 and oboe 11 overlap by 16 bars, and bassoon 7 and oboe 7 by 6 bars.

This structure then forms the basis of the meta-canon: the material for each value of n is the same in both voices, wherever they occur. This means that in sections 11 and 7 the voices really are in canon, and in section 1 they are in rhythmic unison. All of the other sections are not self-contiguous, so the voices perform a canon-at-a-distance. In addition to this basic canonic principle, odd-numbered sections are inverted between oboe and bassoon, while even-numbered sections are performed recte.

There’s not really much in the way of musical example I can post until I’ve explained the derivation of the material (and finished writing it all!), so in the mean time, here’s a colourful illustration of the structure of the canon:

Oboe concerto teaser: I

As I mentioned before, I am working on an oboe concerto. This is a commission for the Handel Collection, and will be performed on 5 July 2011, 13:0014:00 at St Stephen Walbrook (warning: site plays music), 39 Walbrook, London, EC4N 8BN. Please come along if you can!

I’m currently working on the central movement, which is a sort of meta-canonical sarabande. I thought it might be interesting to say a few words about it while it takes shape.

The movement is based around three key ideas:

I’m going to describe the chaconne here; the other elements will come in future posts.

Viola, Violoncello and Double bass staves

One of my favourite musical techniques is the process of gradual transformation found in bellringing. In this process, a row of pitches is repeated over and over, but with each repitition any two neighbouring pitches can ‘change‘, swapping places. In campanology, there is a great discipline to coordinating a group of ringers to perform a whole sequence, or peal, of changes without fault; on a smaller scale this is a great way to produce musical material that maintains unity while developing over time.

In my chaconne, two rows are transformed: the viola has a descending chromatic scale from a′, while the cello and double bass have a figure that opens chromatically like a funnel from A:

On the second repetition, I apply a simple transformation to each row: in the viola, places 1 & 2 swap places, as do 3 & 4, 5 & 6, 7 & 8, 9 & 10, and 11 & 12. In the bass, places 1 and 12 remain in place while 2 & 3, 4 & 5, 6 & 7, 8 & 9, and 10 & 11 all swap:

If I were then to apply the same transformation again, I would get back to where I started, so instead I apply the transformation originally used in the viola to the bass, and vice versa:

If I continued this long enough (24 changes), I would return to where I had started, having performed a plain hunt maximus (ie, the simplest possible pattern on 12 pitches); for this movement, I am more interested in moving from one row to another, rather than returning, so I finish the piece after 13 rows (12 changes), at which point both voices have reversed:

As you will see from the excerpt at the top, I have set these notes to an ostinato rhythm, whose purpose is to support the counterpoint above; each note in the row takes up a bar, so we have a total of 12 × 13 = 156 bars, which, anticipating my next post, happens to be 2 × (1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 + 10 + 11 + 12).