An idea doesn't exist until we become aware of its realizable potential. An idea in itself, musically speaking , does not exist; it is a reaction to what surrounds us culturally. Before actually moving on to the idea-realization pairing in his own field, the composer must first apprehend, grasp this pairing in others - though he can never be sure of grasping it in its full meaning, or even accurately. For once an idea has been realized, it can no longer really be deciphered. If we try to retrace the path from idea to realization, we'll never grasp the deeper motive: it's been burnt up by realization; it's been burnt up, so it's disappeared as such to become a work of art. A real work is the annihilation of the work's original desire; it is both the overcoming and the negation of the original idea.

All we can do is analyze it, trying to reconstitute in reverse the path from realization to idea: a kind of perversion of the work through investigation. Often, analysis simply transcribes the musical notation of the work using a different vocabulary and notation. This transcription, using a different code, gives a precise account of the finished object and its slightest accidents, but does not focus on the why of this object; for other criteria come into play that escape formal analysis, being in the realm of intuition, of the irrational. Analysis is therefore the pursuit - vain, no doubt, in the absolute - of the labyrinth that joins idea to realization. However, in any important work, the realization is full of accidents, in relation to the initial idea, accidents that make the realization more interesting than the idea. Analysis cannot hope to reconstitute the labyrinth, recreate the accidents, resume the proliferation. Non-creative analysis mutilates; it leads to a comfortable description, whereas doubt and dissatisfaction should be its hallmarks.

The most seductive situation is to create a labyrinth from another labyrinth, to superimpose one's own labyrinth onto that of the composer: not to try in vain to reconstitute his approach, but to create, from the uncertain image we may have of him, another approach. Productive analysis is probably, in the most casual case, false analysis , finding in the work not a general truth, but a particular, transitory, subjective truth.

The critical casualness of this intuition is not immediately apparent. When the composer's activism is stronger than the personal quality of his or her invention, the intuition vis-à-vis the works of the past is only weakly effective, because the invention creates precisely the detonation from a fragmentary, intuitive analysis. There's certainly an apprenticeship to be made, of seizure: learning to invent according to the work of others. This development involves grasping oneself no less than grasping history. It's a question of knowing how to place oneself in a territory whose coordinates one gradually establishes; more than the obviousness of a "geographical" situation, it's necessary to understand the necessity of insertion. We are therefore far removed from the restricted formalism of analysis; we constantly refer to the inescapability of the work.

We tend to believe that by observing a model, by analyzing an exemplary object, the experience of others can be transmitted and absorbed in its totality. What's more, the model's advantages will be consecrated and magnified, while its weaknesses will be recognized, analyzed and rejected. These sequences would be far too simple. Curiously, even the most gifted personalities do not always make up for deficiencies, which remain a permanent shortcoming that no amount of confrontation can remove. The gift, oriented in a very individual way towards such specificities of language, acquires skill, virtuosity in one direction, and seems incapable, in another direction, of manipulating, transforming the elements supplied to it.

What, in the final analysis, is the craft that is transmitted in such a random and precarious way? And how does invention relate to craft in its early stages? Confrontation with one's predecessor yields results that are far from predictable, and are the result of the non-coincidence of gesture between the observed and the observer.

Classical musicology has for the most part been concerned with describing the evolution of a personality by describing certain procedures of imitation to arrive at the most absolute marks of originality. What she fails to describe, and what would be by far the most interesting, is the permanence of a given composer's gesture : firstly, how this gesture is applied to others, thus deforming the gesture of others by adding or subtracting in a specific way; secondly, how the gesture becomes established, recognizing itself as such; thirdly, how it becomes refined, making itself absolutely irreducible to any other gesture.

The craft is acquired and not acquired in relation to one's predecessors; there are innate dispositions that will make one choose this or that characteristic of the language. The transmission of the craft is therefore extremely random, and depends above all on a personal profile that gradually takes shape and takes shape through privileged encounters, where choice and chance share the responsibility. There is no valid global solution to these questions of profession; only the reactions of the individual will create for himself a kind of provisional law. What is it, then, that makes it possible to recognize craft even in the most adventurous, insecure works? And even then, we only seem to be talking about music that is, if only remotely, part of the tradition and evolution of written music, music in which transcription through writing plays a role that is not essential, but indispensable. But what can we say about music invented with newly acquired technology, where invention can transgress the limits of transcription, where we more often than not note actions likely to produce different, unforeseen results and sound objects? Increasingly, the notion of craft is difficult to grasp, and therefore increasingly difficult to transmit.

Craft is a multifaceted concept that mediates the composer's gesture through invention. It is, first and foremost, predictive: anticipating writing, formal work and pragmatic solutions is the part of the craft that can be learned. The craft relies no less on deduction, which helps us to spot the idea, then to make it proliferate. Invention is not a profusion of deduction. In this sense, craft is the respondent, the correspondent of invention. Contrary to the apparently restrictive nature of the term, deduction is not simply the rational application of an ability to draw consequences from a given fact; deduction can be highly illogical or irrational, and its force can short-circuit the chain of consequences from one step to the next. Deduction is implicit in the Idea, the material perfection of the Idea being subordinated to the powers of deduction.

The craft is both foresight and transgression; it can be the fundamental hinge of invention if the transgression is made from knowledge, creating a responsibility as yet unexplored in its forms and consequences. The profession of the existing constantly refers back to the profession of the imagined, so that the unknown emerges from the known, so that this chain of events is ineluctable and irreversible.

P. B.