Musical scores are information devices that allow a composer to lay down certain parameters of a performance. In practice, scores define what is considered context-independent by the composer or by a certain cultural tradition. No score will thus ever totally determine all aspects of a musical performance: some elements of music making will always be contingent - and thus improvised. In fact, there is no clear line between improvised and composed music – hence my term for score-based music with elements of improvisation: comprovisation. I use this term for all music that draws not only on the contingent moment of performance but also on context-independent rule systems or scores.

In my recent comprovisation scores for larger ensembles (Inside a Native Land [Berlin 2005], Vineland Stelae [Los Angeles 2007], Racines Ephémères [Vienna 2008], Transience [Nürnberg2009/Malmö 2011], and jaali [Stuttgart 2011]) I have explored the combination of improvised, comprovised, and composed elements of a work not only in structural segmentation but also in layerings of sub-ensembles and in what I call "encapsulated traditions".

Just as no marriage can be reduced to the enactment of a written contract, no music is ever completely fixed - and there is no such thing as free improvisation much in the same way as there is no such thing as entirely free love. Music, especially improvised music, is a quintessentially polythetic art i.e. it cannot be understood in any convenient, conceptual overview – it must be performed one event, one moment at the time. 

Thus, even the most fixed acousmatic tape music has depended for its spread on the vagaries of funding, performance space, technology development - and the contingent elements of each single, unique performance. This is why the "diffuseur sonore" has slowly gained acceptance as a special type of musical performer - and pedagogical programs to teach, think and critically examine sound diffusion are currently coming into being.

At the other extreme, it is not by chance that the ideology of free improvisation appears at the same historical moment as does the ideology of free love. But, while the latter was soon deconstructed as a naïve and often oppressive strategy in the gender wars, the incantations promoting a truly free music unfettered by rules and regulations, celebrating the incommensurable beauty of the present moment, still haunt the program brochures and the apologetic texts of its proponents. 

Every performance can be said to consist of two types of elements: those that stay more or less the same between different performances,  and those that are unique to the context of this particular performance.   

The first - let us call them “context-independent” or “repeatable” - are what people usually refer to when they speak of an “artistic work” or a “composition.” The second type of elements - called “contingent” here - also has a wide range of names, such as “chance procedure,” “playing it by ear,” “inspiration,” “arrangement,” “improvisation” or “interpretation” and many more. 

Notations and scores are among the most important conceptual devices used to distinguish between some of the contingent and context-independent elements of a performance.  For example, whereas eurological  art music traditions considered it important to notate pitches and their associated durations, Chinese (or sinological) art music  traditions considered pitches and their associated playing technique to be the ideal pair of notated parameters, leaving the aesthetic significance and practice of duration largely to oral transmission and to the individual musician’s “touch”- similar to the view that eurological art music took towards articulation until the mid-20th century. 

In addition, many traditions closely couple certain performance parameters to fixed combinations, thereby prompting the development of bi- or multivalent notational "objects" : in this view, each possible traditional western basic "note" consists of an unique and graphically unified combination of two differently graded series of symbols that taken together signify both pitch and duration of a musical event.

No score will thus ever totally determine all aspects of a musical performance: Music inventors and composers must adhere to or establish notational conventions that enable them to control certain parametes of music making - but, in this process, the very perspective of notation that they adopt in order to control has side effects: it creates a free space, a musical realm beyond their control. Some elements of music making will always be contingent - and thus improvised. But the opposite is also true - in creating music on the spur of a moment improvisers are also establishing all kinds of momentary scores - and the word "score" here explicitly includes all kinds of rule systems, embodied reflexes and inner representations that prompt a performers to play the next note…

In fact, there is no clear line between improvised and composed music – hence my preference for another term: comprovisation.

Comprovisation scores abound in Jazz and in New Music since the 1940s. Three main dramaturgical-architectonical-textural (fused as: "architexture") models appear to overwhelmingly dominate traditional - and large territories of contemporary - comprovisational practice: 

1.) dramaturgically "linear" variations on a stable melodic-harmonic formula or on a rhythmic cycle – all dramaturgical developments are either of a transient nature, to be resolved in regulary recurring cadential moments - or they are placed within a "non-surprising" framework, e.g. a strict monotonic accelerando structure, as in the case of North Indian art music.

2.) "free" forms of organic growth and/or decay: isolated beginnings, slow congealing into coherence, exponential growth of density ending in a "tout-son" situation where individual contributions dissolve in the overall sound. As musicians cannot hear each other, they drop out: texture and density lighten again. At this point the others resume - and a surge to next peak follows - etc.

3.) ritualized improvisations with composed section markers, or compositions with embedded improvisations – in both certain blocks of time are set aside for improvisation [whether with an elaborated groove or drone as in Jazz and many modal improvisations, or without such a listener's lifeline, as in Lutoslawski scores]. Such cadenza-like "absences" of the composer are like a carnival of inverted roles between composer and performers. 

In "Amadeus" (the film), Peter Shaffer lets Mozart say: "In a play, if more than one person speaks at the same time, it's just noise.  No one can understand a word.  But with music, with music you can have twenty individuals all talking at once, and it's not noise - it's a perfect harmony. Isn't that marvelous?" A praise of polyphony in a movie! But Mozart, except for some of his last works, is not considered to be a particularly polyphonic composer. What exactly is he talking about ?

There are two basic types of polyphony – one concerned with the unfolding of and reflections on a central idea kernel: this type could be called exegetical polyphony. It can be found in bicinia, canons, ricercars, cantus firmus and chorale variations, fugues etc. In all these forms, new musical realities are unfolded from or reflected in one source rather than composited from different sources.

The other type of polyphony is about creating mutually illuminating relations between previously unrelated ideas. This is a much rarer type of polyphony – it can be heard primarily in isorhythmic motets, in double and triple fugues, in some developmental sections of classical sonatas etc. One could call it relational polyphony. 

Interestingly, whenever Mozart used polyphony, he tended to use it in this relational mode: in the last movement of his last symphony four different themes jostle each other and through their constantly changing relation to each other create a kind of structural harmony that is very different from that provided by exegesis.

The idea of relational polyphony is at the heart of my recent approach to comprovisation - how could one imagine a music for 5 – 40 or more improvising musicians that satisifies the refined taste for dramaturgical and textural differentia¬tion most western art music listeners have acquired by listening to written music? 

In my view, every improvisation style is essentially traditional in nature, as it de¬pends on the oral and enacted transmission and conventionality of complex knowledge structures: Each improvisation style creates a rich cloud of determinants and constraints that must be enacted by many different creators.  An improvisation tradition thus is an organism, in itself consisting of many suborganisms - down to essential rhythmic, gestural, and textural cells.

Since 1996, I have conceived a number of comprovisation scores for larger ensem¬bles, and since ca. 2003 they are based on encapsulated traditions: The musicians are not given explicit music, but rather blueprints on how to improvise within a cer¬tain framework. Performers are given precise constraints and instruction, which they need to learn by heart and embody. These instructions can be very detailed, and are often illustrated by a notated example of this invented tradition. Rehearsing  such a score is different from learning a written score, but also from learning how to improvise a new song in one's own style: this approach is almost like learning a North Indian raag. 

The next phase then is to create a polyphonic score that would allow to superimpose these different encapsulated traditions within a convincing architexture. Each of my works adopts a different strategy for this step. Two examples:

In RACINES ÉPHÉMÈRES (2008) for 8 wandering musicians, obbligato conductor and live-electronics, the architexture leans heavily on the principle of the variation: In each of the 64 variations on a theme by Claude Vivier, a different selection from the 9 encapsulated traditions present in this work is superimposed in a unique way. Each variation includes precise instructions on how musicians interact with other musicians as they change places and are re-configured throughout the piece.

There is no music without some kind of score - not necessarily written out, but as a mental architexture necessarily present in the context of music creation and mak¬ing: scores, as Lawrence Halprin understood them, "orchestrate design, participa¬tion, events and activities that visibly delineate, generate, and sustain a project": this definition makes no marked distinction between the sonic and the meta-sonic as¬pects of music making. If hardly any written and performed score or acousmatic soundfile functions purely on a sonic level, improvisations definitely cannot be un¬derstood without the context that shapes music making – what Christopher Small has called "musicking".

Traditions, contrary to their reputation in new and improv music circles, are not old and unoriginal ways of making music, they are an innate (but mostly unacknowl¬edged) part of every score: without a tradition, scores become unreadable objects.