Dome Songs is a song cycle for a chamber ensemble of baritone, flute (doubling on alto flute), bass clarinet, bassoon, cello, and piano. The work's text is derived from two lines of Percy Bysshe Shelley's elegy Adonaïs, quoted in context below. The lines used in Dome Songs have been italicized: The One remains, the many change and pass; Heaven's light forever shines, Earth's shadows fly; Life, like a dome of many-coloured glass, Stains the white radiance of Eternity, Until Death tramples it to fragments.--Die, If thou wouldst be with that which thou dost seek! Follow where all is fled!--Rome's azure sky, Flowers, ruins, statues, music, words, are weak The glory they transfuse with fitting truth to speak. The first song employs the text in its original form, and the next two rearrange and contort these lines to create new meanings and emphases, altering the positions and roles of key words in the passage. For example, the order of "life" and "eternity" reverse from the first song to the last; the adjective "many-coloured" refers to a different noun in each case; and the role of "stains" changes in each song, altering the rhythm and meaning of the text. Local and large-scale text painting is employed, principally in the incorporation of many simultaneously active arch-like structures governing most compositional parameters. The texts of the songs are as follows: I. Life, like a dome of many-coloured glass, Stains the white radiance of Eternity II. Glass-like life of many-coloured stains; Eternity, a radiance: the white dome III. Eternity, the many-coloured radiance, Stains a white dome of glass: life. The pitch material is derived principally from 12-tone techniques, but secondarily through intervallically controlled free atonality. The principal melodic lines are drawn from 12-tone rows designed to exhibit several levels of symmetry and a carefully calculated internal structure. The second song, for example, employs this row: E 8 0 6 3 4 T 9 7 1 2 5. This row's disjoint segmental trichords form a symmetrical structure ([014] [013] [013] [014]), as do its disjoint segmental tetrachords ([0146] [0167] [0146]. Additionally, the interval between the last pitch class of each consecutive trichord and the first pitch class of the next is always a tritone. These traits are summarized in the chart below: E 8 0 6 3 4 T 9 7 1 2 5 Although each song uses a different row with different characteristics, each row is designed using similar principles, resulting in a cogent, musical flow from one song to another. The symmetrical properties of these rows are exploited compositionally in each song. Harmonic and contrapuntal material generally derives from the row itself, but occasionally includes intervallically generated material designed to complement the 12-tone row of principal importance. The first song, for example, uses two row forms (one pitch row and one pitch-class row, explained below) that function together at any given time along with accompanimental voices that are freely composed. The row is embellished with a single suspension figure in the melody, which is added to emphasize the midpoint of the text and provide a half-cadential feel. In contrast, the third song consists almost entirely of precise row forms, deployed in canon. Instrumentation is controlled by unique but related techniques in each song. The first song, for example, controls instrumentation as a formal element, associating different instruments and textures with each part of the form. The bass clarinet is restricted to the first half of the song, and the piano to the second; meanwhile, the flute and bassoon alternate in an order related to the row forms that control pitch-class content. The second song, however, employs row transformations that dictate exactly which instrument plays each note. The third song combines these two approaches, using a single instrumental row to prescribe the formal roles of each instrument across the entire song. Instruments are assigned numbers (0 - bassoon, 1 - bass clarinet, 2 - baritone, 3 - flute, 4 - piano) and a five-note row is chosen (43102). Two forms of the instrumental row (one prime and one retrograde) function simultaneously, as below, resulting in a central duet between voice and piano: 4 3 1 0 2 2 0 1 3 4 2 0 1 3 4 4 3 1 0 2 Register is structurally essential in every movement and selected by a number of criteria, which vary among from one song to the next. The first song, for example, employs both a pitch-class transform and a pitch transform at any given time, based on different inversions or transpositions of a common row. The pitch-class row will determine the ordering of the melody (which sometimes weaves among instruments and is not solely the domain of the voice), while the pitch set determines the register (if the C in the pitch set is C4, the C in the corresponding melodic row will be C4 even though C's positions in the two rows are usually different). The first pitch-class row, for example, is T8 (4 5 8 E T 3 2 7 6 9 0 1), while the first registral row is T0 (-4 -3 0 -9 -22 7 6 -1 -2 1 4 5). The composite pitch row, then, which retains the pitch class ordering of T8 but uses the registers of the pitches in T0, is (4 5 -4 -1 -2 -9 -22 7 8 -3 0 1). This technique is used throughout the song. In the second song, register is carefully employed to emphasize particular intervals in the pitch row and create large-scale form; for example, the boundary interval of any given disjoint segmental pitch hexachord is always interval class 2 in the A and A' sections, but interval class 6 in the B section. Each pitch appears only once in the underlying row forms (with a single exception, made out of compositional necessity because of instrumental range limitations). For example, after an A3 appears as part of a row form, that pitch does not reappear in any later row form in the movement; any later A that is part of the row must occur in a different register. Especially in the second song, registral expansion and contraction emphasize formal subdivisions. Compositional methods applied to rhythmic control vary from song to song, but are always based on contrapuntal relationships rather than strict durational rules. The second movement is the most rigid in its serial control of rhythmic elements, employing various transformations of a rhythmic row that dictates neither time-point nor duration, but rather the grouping and subdivision of each series of notes, particularly those in the instrumental parts. This system allows a measure of compositional flexibility while imparting meaningful structure to the piece. For a simple example of this, see the beginning of the second song: The rhythmic row operating here is 2 1 3 0 5 4 5 0 4 1 2 3. I add 1 to each row element in order to derive note groupings, from single notes to sets of six, with one exception. For this song I allow two different interpretations of a '5' in the row. While a '5' is often realized as a set of six attacks (as in measure 4, alto flute), it is sometimes also realized as a set of three fast notes, as in the piano and bassoon above. This allows for a more musical introduction to the movement. The third song combines the rhythmic techniques of the first and second by using a row to indicate the most important rhythmic groupings in each section of the form. The rhythmic row is simply T0 mod 6, resulting in the following: Pitch Row (T0): 4 3 1 0 2 5 6 9 E T 7 8 Rhythm Row (T0 mod 6): 4 3 1 0 2 5 0 3 5 4 1 2 Row elements are deployed in pairs beginning at the structural downbeat (measure 5), resulting in initial groupings of five and four notes. The central piano/vocal section uses its own instance of the row. The rhythmic structure of the piece is summarized below. Measure 1 5 17 24 34 Section Prelude A B A' Coda Rhythmic Row 1 43 10 25 03 54 12 Rhythmic Row 2 431025035412 This compositional system is highly controlled yet allows a measure of creative freedom, resulting in a lyrical, expressive, yet strongly and compellingly structured piece of music.