Speak the Middle Tongue – Take the Forked Road: A Theory of the Voice
The voice is the sign of pain and pleasure…
Aristotle, Politics, 123a, 10-18.
The grain of the voice, Barthes tell us, “is… the materiality of the body speaking its mother tongue; perhaps the letter, almost certainly significance.”  To comprehend this definition, we must rigorously avoid a misunderstanding easily succumbed to; that to signify something is to communicate it. To ward off that misapprehension we must answer the following questions: How are we to understand materiality and significance, and their relation? What does the italicization of significance connote? Barthes’ use of materiality is straightforward; it refers to “the sonic effects of the tongue, the glottis, the teeth, the mucous membranes, the nose;”  to the “body in the voice as it sings, the hand as it writes, the limb as it performs.”  These embodied forces determine the diction of enunciation, which constitutes its ‘grain,’ that allows us to recognize the identity of a speaker when he/she speaks. We recognize voices as we recognize faces – through their embodiments.
Barthes theoretically allies the grain with the geno-song, a biological, materialist concept he transposes to music from Kristeva’s linguistic analog – the geno-text – individual works of pheno- or species-text exemplified by genres like romance or science fiction texts. The geno-song is defined thus:
[it] is the volume of the singing and speaking voice, the space where significations germinate ‘from within language and in its very materiality’; it forms a signifying play having nothing to do with communication, representation (of feelings), expression; it is that apex (or that depth) of production where the melody really works at the language – not at what it says, but the voluptuousness of its sound-signifiers, or its letters – where melody explores how the language works and identifies with that work. 
To illustrate the materiality of the geno-song, I’ll play a clip of the Tuva throat singer, Borbannadir.
Significance, then, is italicized to warn us against mistaking it for communication, representation, or expression. It is what works at language but not at its meaning. Meaning, for Barthes, is a product of the reductive forces of the pheno-text by which culture enforces limits to understanding, to significance, by “reconcil[ing] the subject to what in music can be said: what is said about it, predicatively, by Institution, Criticism, Opinion,” by which he means the codes of langue that always precede and police the voice.  Significance of the grain, Barthes acknowledges in a passing parenthetical allusion to another of his foundational texts, derives its value when the text emerges in the work.
In order for a voice to have grain, to have non-communicative significance, it must break the codes of works, of pheno-songs and pheno-texts, and emerge into the text of the geno-song through listening to the relation of the body of the speaker, singer, or player. The relation, he tells us, also transcends individualistic subjectivity because it is erotic and physiological – it is not a psychological subject who sings or listens, the voice is not an expression of any subject, but it’s dissolution, produced outside of the laws of culture, beyond the valuations of ‘I like’ or ‘I don’t like.’ 
We are all familiar with these poststructuralist themes: text vs work, authorial death, corporeal jouissance, the politics of language. My intention here is to remind us of what the stakes were then, (and which I strongly believe are still relevant today), in putting significance beyond reductive codes that normalize meaning, and in relation to a decidedly non-subjective but still embodied materiality. Sound, music, voice, all have the critical capacity to de-subjectivize, to break free of the narrow circuits through which the codes of culture restrict most of us to ‘expressing’ only individuality. The politics of the grain of the voice aims to produce collective responses against liberal humanism’s ‘voice’ that atomizes the social through its infinite and repetitive broadcast of bankrupt individualist, subjectivist fantasies. While this claim may appear hyperbolic, it is what Barthes intended, and accurately represents the political aims of much poststructuralist thought – to free our voices from a double dominance; from the dominance of normalizing pheno-songs on the one hand, and on the other, from communications of reductive meaning, between singers and listeners.
I will now turn to Derrida’s analysis of ‘monolingualism’ as a means by which to define voice, not as a work but, by analogy, as a ‘text.’
Monolingualism opens by staging the scene of a dialogue about “performative contradictions.” The umbrella formulation of the contradiction reads: “I have only one language; it is not mine.”  The contradiction, thus formulated, may easily be justified. In Kristeva’s terms, the contradiction lies between pheno-text and geno-text; for Barthes analysis of the voice, between pheno-song and geno-song. I propose here another dichotomy between pheno-voice and geno-voice, which I will formulate in a moment. Each of these oppositions is a species of the Sassurian legacy; parole refers to having one language through one’s ability to speak it; while langue refers to the impossibility of ever possessing a language at all because it forever supersedes the capacity of speech. A second defense of Derrida’s performative contradiction lies in the dialogic; in the reciprocal, bi-directionality of speaking/listening, in the inevitability of networks of co-interpolations; signifying networks in which speakers call listeners into being, and vice versa. Recognizing that the stakes of ‘giving voice to’ involve communication, but not necessarily significance – what is it that is given voice, for what purpose, and by what means? – we quickly find that a long series of antinomies unfolds here: truth and lie, confession and judgment, proof and construction, monologue and dialogue. This partial series demarks some of the undecideable parameters of “performative contradiction,” and therefore of what the act of giving voice signifies.
This series of antinomies also allows us to suggest initial definitions of pheno- and geno-voices.  The pheno-voice is an agent of communication whose role is to negotiate contradictions in the act of giving voice, in its performance of irresolvable contradictions whose function is not to generate, but to suspend, meaning.  The geno-voice arises in a particular network of interpolations as an advocate for some vector of signification; it is always a product of objective conditions both exogenous and endogenous to it.  This means that the geno-voice is an assemblage of all other voices (the text emerging in a work) that have responded to the call of a given performative contradiction.
A voice is an agency that advocates by calling for a particular significance for a given network of co-interpolations. It is a polygenetic agency, a call without a singular agent, without identity – polyphonic, polyglot, and polyatomic – a complex network of nodes and characteristics resonating along the vocal scale of the network of terms show here on the paradigmatic axis.
To put this in the precise language of performative contradiction, a geno-voice is not more objective than subjective. In Barthes’ terms paraphrased above: it is produced by listening to the relation of the body of the speaker, singer, or player; with the additional qualification of plurality or relations among speakers, singers, and players. A voice interpolates values, and is therefore essentially, necessarily a sociopolitical agency called into being through its encounters with antinomies, with the Other that is both exogenous and endogenous to it. To recognize that we BOTH always and never speak only one language, that language is BOTH ours and not ours, is to know, as Gevirtz has so beautifully put it, that when we give voice to anything at all, it is to speak with the middle tongue and to take the forked road.
I now turn to some examples to illustrate this theory of the voice. In 1979, the Caribbean poet Edward Kamau Brathwaite gave a remarkable lecture at Harvard entitled, History of the Voice, published in 1980 in book form. The aim of this work was to show how Black Caribbean poetry emerged simultaneously with and through the musical geno-songs of Calypso and Reggae with the conscious intent of distinguishing itself from the dominance of British English taught in Jamaican schools. His lecture demonstrates in extraordinary detail how these geno-songs are based in the materiality of the Caribbean environment, like its weather, and from the polyglot patois of black Jamaican idiom, in order to resist the master’s poetic language exemplified by the Shakespearian metrical model of iambic pentameter.
I’ll cite here only one of his many examples, in which he compares Shakespeare to his own poetry, to illustrate this.
(IP) To be or not to be, that is the question.
(Kaiso) The stone had skidded arc’d and bloomed into islands
Cuba San Domingo
Jamaica Puerto Rico
Brathwaite, ‘Caribbean theme: a calypso, CQ4: 3&4 (1956) p. 246;
Rights of Passage (1967: as “Calypso’), p. 48; sung by the author on Argo DA101 (1969), PLP1110(1972).
He describe the difference between these lines in this way:
…not only is there a difference in syllabic or stress pattern, there is an important difference in shape of intonation. In the Shakespeare…, the voice travels in a single forward plane towards the horizon of its end. In the kaiso, after the skimming movement of the first line, we have a distinct variation. The voice dips and deepens to describe an intervallic pattern.
Barthe’s concept of the geno-song is readily apparent here; the “shapes of intonation” the dipping and deepening of the voice, and the emphasis on the “intervallic pattern,” are all terms which stress embodiment. Based on this type of polygenetic, poetic analysis, he then elaborates the concept of “nation language:”
Significance, meaning, is co-created in the intervallic spacetime between the material recitation of spoken/sung sound and song; it arises as much from noise as from written language. But nation language also demonstrates the polygentic, performances of contradictions that are the fundamental condition of monolingualism, what Brathwaite calls the ‘total expression” that determines the uniqueness of Caribbean poetry, that is based on
The voice speaks to its own mouth
and also from a speech external.
Prosthesis, Susan Gevirtz
Total expression is performed in the antinomic space between listener and singer, between the voice and the call; that is to say, the singer is also and simultaneously a listener just as the listener is simultaneously a singer. Singer and listener reciprocally interpolate one another, but do so relative to the historically unique immanence of breath, poverty, and Jamaica’s material environment. It is the particular material conditions of its island state that calls forth the particular vector of significance as both noise and language, and that gives polygentic voice to the particular Caribbean agency without agent of nation language. This polygentic voice is a corporeal agency that embodies a diction recognizable as what Bourdieu has called the collective-individual, and Spivak has called the planetary subject – subjects that resist social atomization because the listen to the call of total expression, and learn to give it voice to the network of co-interpolations.