I have always held a deep reverence for the mountains. The sense of oneness with the universe and the power of creation, the awareness of space, the purity of the air and the inexorable force of all the elements can often be overwhelming. It is hardly surprising that so many of the great Biblical and mythological dramas are set there - the Transfiguration, the Psalms, the Bacchae.... And yet with all the sense of vastness comes also an acute sense of identity - each bird, each animal, each stone and each stream has its own spirit, and its own particular place in the cosmos. Communication between all these spirits seems constantly alive and immediate.
In seeking to explore these phenomena in Mountain Language I decided to base as many elements as possible on natural models - these include all the elements of rhythm, melody, harmony, intensity, texture and form, the especially crucial element of spatialisation - and, of course, sound itself.
All the sounds in Mountain Language are derived from sounds which are indigenous to the mountains, and which are designed to transmit signals over large distances - cowbells, church bells, alphorn, and the elemental sounds of wind and rain. These sounds are deployed by the two live performers (alphorn and cowbells), two samplers (triggered by the cowbells themselves) and a computer, triggered by a keyboard player who acts as a conductor/co-ordinator. Thus the 'language' implied by the title does not merely refer to the visible dialogue between the live performers, but much more to the complex network of signals, responses, echoes and other dramatic interactions between the live musicians and the multitude of invisible, spatialised sounds.
When, at the very outset, I started to consider the basic harmonic language of the piece, I was immediately faced with the problem that the live alphorn, magnificent instrument though it is, is limited to more or less the first twelve partials of the harmonic series of a single fundamental (F). In acknowledging this limitation, the harmonic structure of Mountain Language is based on varying combinations of the spectra of the three main sound sources (alphorn, cowbell, church bell), with each spectrum based on a fundamental of F and encompassing some 70 partials. Since each of these spectra is totally different, it was possible to create a harmonic language of great richness and versatility, notwithstanding the fact that no transposition other than octave transposition would be employed.
The melodic models in Mountain Language are from birdsong. In analysing over a hundred birdsongs for their melodic, rhythmic and intensity data it was possible to apply this data to both the sounds and the harmonic scheme already described. Once the melodic data is transferred to a particular spectrum (or combinations of spectra), melodies can be transposed 'by degree' so that they always articulate pitches within the same basic harmony.
They can also, of course, be sped up or slowed down to suit different musical contexts.
The most obvious (and most literal) examples of birdsong models can be heard in the extended passages triggered by the cowbells, where the birdsong fragments serve as highly elaborate ornamentation to the cowbells. These passages reveal not only the birdsong's wellknown melodic and rhythmic detail, but also an extraordinarily sophisticated play of intensities, or dynamics.
But the models are used also for all the alphorn material. By taking a small fragment (even two or three notes) of birdsong, slowing it down, applying it to the alphorn spectrum and then transposing it down the degrees of that spectrum, the song begins to resemble a typical alpine horn call. Thus whilst the horn calls heard at the opening, for instance, may sound like archetypal horn calls, as they climb the spectrum, getting longer and faster, they gradually reveal themselves to be fragments of birdsong. In this way all the melodic material is unified by a common model, and can move up and down in register without technically 'transposing'.
A similar process is used to realise the models of contour, texture and intensity in wind and rain. In the case of the wind, this in fact creates a kind of aeolian harp (heard, for example, near the beginning together with the very same wind sound that created it). In another passage the rain textures are used to generate a transformation from rain to birdsong-like textures.
The spatialisation in Mountain Language has been modelled on the geographical location of eighty-nine peaks surrounding a central reference point, the village of Ishgl, Paznauntal, Tirol. These eighty-nine points provide 'platforms' for the work's principal polyphonic discourse of 'Responses' and 'Echoes', as well as for the many elemental sounds. In addition to the eighty-nine peaks, there are three other important locations following the line of the valley - these are the churches of three neighboring villages - that of Ishgl in the centre, and those of Kappl and Galtuxxr at opposite perimeters of the area.
The form of Mountain Language follows the progress of a single day-cycle which could roughly be described as follows: Night - dawn - church bells announce the morning Mass dawn chorus - morning - light rain eventually develops into a storm in the afternoon much activity in the late afternoon - evening chorus - dusk - church bells announce Evening Prayer - sunset - night.
(c)James Wood, May 1998