Borrowing Julia Kristeva’s terms ‘phenotext’ and ‘genotext’ (where the former serves to communicate competently, while the latter is a process that articulates ephemeral or non-signifying structures), Roland Barthes’s essay identifies and examines the split between voice and language. He heralds the grain of the voice, or ‘genosong’, as a form of bodily communication that circumvents the laws and limits of the linguistic sphere, and reveals the materiality of language from within.
Language, according to Benveniste, is the only semiotic system capable of interpreting another semiotic system (though undoubtedly there exist limit works in the course of which a system feigns self-interpretation — The Art of the Fugue). How, then, does language manage when it has to interpret music? Alas, it seems, very badly. If one looks at the normal practice of music criticism (or, which is often the same thing, of conversations “on” music), it can readily be seen that a work (or its performance) is only ever translated into the poorest of linguistic categories: the adjective. Music, by natural bent, is that which at once receives an adjective.
The adjective is inevitable: this music is this, this execution is that. No doubt the moment we turn an art into a subject (for an article, for a conversation) there is nothing left but to give it predicates; in the case of music, however, such predication unfailingly takes the most facile and trivial form, that of the epithet. Naturally, this epithet, to which we are constantly led by weakness or fascination (little parlour game: talk about a piece of music without using a single adjective), has an economic function: the predicate is always the bulwark with which the subject’s imaginary protects itself from the loss which threatens it. The man who provides himself or is provided with an adjective is now hurt, now pleased, but always constituted. There is an imaginary in music whose function is to reassure, to constitute the subject hearing it (would it be that music is dangerous—the old Platonic idea? that music is an access to puissance, to loss, as numerous ethnographic and popular examples would tend to show?) and this imaginary immediately comes to language via the adjective. A historical dossier ought to be assembled here, for adjectival criticism (or predicative interpretation) has taken on over the centuries certain institutional aspects. The musical adjective becomes legal whenever an ethos of music is postulated, each time, that is, that music is attributed a regular—natural or magical—mode of signification. Thus with the ancient Greeks, for whom it was the musical language (and not the contingent work) in its denotative structure which was immediately adjectival, each mode being linked to a coded expression (rude, austere, proud, virile, solemn, majestic, warlike, educative, noble, sumptuous, doleful, modest, dissolute, voluptuous); thus with the Romantics, from Schumann to Debussy, who substitute for, or add to, the simple indication of tempo (allegro, presto, andante) poetic, emotive predicates which are increasingly refined and which are given in the national language so as to diminish the mark of the code and develop the “free” character of the predication (sehr kräftig, sehr präcis, spirituel discret, etc.).
Are we condemned to the adjective? Are we reduced to the dilemma of either the predicable or the ineffable? To ascertain whether there are (verbal) means for talking about music without adjectives, it would be necessary to look at more or less the whole of music criticism, something which I believe has never been done and which, nevertheless, I have neither the intention nor the means of doing here. This much, however, can be said: it is not by struggling against the adjective (diverting the adjective you find on the tip of the tongue toward some substantive or verbal periphrasis) that one stands a chance of exorcising music commentary and liberating it from the fatality of predication; rather than trying to change directly the language on music, it would be better to change the musical object itself, as it presents itself to discourse, better to alter its level of perception or intellection, to displace the fringe of contact between music and language.
It is this displacement that I want to outline, not with regard to the whole of music but simply to a part of vocal music (Lied or mélodie): the very precise space (genre) of the encounter between a language and a voice. I shall straight away give a name to this signifier at the level of which, I believe, the temptation of ethos can be liquidated (and thus the adjective banished): the grain, the grain of the voice when the latter is in a dual posture, a dual production—of language and of music.