266 / R O L A N D B A R T H E S

be exclusively a studio, a workshop, an atelier from which

nothing—no dream, no image-repertoire, in a word no "soul"—
will overflow and where all musical doing will be absorbed into
a praxis with nothing left over. It is this utopia which a certain
Beethoven, one not played, teaches us to formulate—whereby
it is possible to foresee in him a musician still to come.
The Grain of
1970

the Voice

Language, according to Benveniste, is the only semiotic sys-

tem capable of interpreting another semiotic system (though
there are doubtless certain limit-works, in which a system
feigns self-interpretation: The Art of the Fugue). How then
does language manage, when it must interpret music? Alas,
badly—very badly, it seems. If we examine the current practice
of music criticism (or of conversations "on" music: often the
same thing), we see that the work (or its performance) is in-
variably translated into the poorest linguistic category: the ad-
jective. Music is, by a natural inclination, what immediately
receives an adjective. The adjective is inevitable: this music is
this, that execution is that. No doubt, once we make an art
into a subject (of an article, a conversation), there is nothing
left for us to do but "predicate" it; but in the case of music,
this predication inevitably takes the most facile and trivial
form: the epithet. Of course this epithet, to which we turn
and return out of weakness or fascination (parlor game: dis-
cuss a piece of music without using a single adjective), has an
 268 / ROLAND BARTHES The Grain of the Voice / 269

economic function: the predicate is always the rampart by of doing here. What we can say is this: it is not by struggling
which the subject's image-repertoire protects itself against the against the adjective (shifting this adjective that comes to the
loss that threatens it: the man who furnishes himself or is tip of our tongue toward some substantive or verbal peri-
furnished with an adjective is sometimes wounded, sometimes phrasis) that we are likely to exorcise musical commentary
pleased, but always constituted; music has an image-repertoire and to liberate it from the predicative fatality; rather than try-
whose function is to reassure, to constitute the subject, who ing to change directly the language used about music, it would
hears it (would this be because music is dangerous—an old be better to change the musical object itself, as it presents it-
Platonic notion? Leading to ecstasy, to loss, as many examples self to speech: to modify its level of perception or of intellec-
from ethnography and popular culture would tend to show?), tion: to shift the fringe of contact between music and language.
and this image-repertoire immediately comes to language by It is this shift that I should like to sketch here, not with
the adjective. A historical dossier should be compiled here, regard to all music, but only with regard to a portion of vocal
for adjectival criticism (or predicative interpretation) has music (art song, lied, or melodie); a very specific space (genre)
assumed, down through the ages, certain institutional aspects: in which a language encounters a yoke. I shall immediately
the musical adjective becomes somehow legal whenever an give a name to this signifier on the level of which, I believe,
ethos of music is postulated, i.e., whenever a regular (natural the temptation of ethos can be liquidated—and the adjective
or magical) mode of signification is attributed to music: therefore dismissed: this name will be the grain: the grain of
among the ancient Greeks, for whom it was the musical the voice, when the voice is in a double posture, a double pro-
language (and not the contingent work), in its denotative duction: of language and of music.
structure, which was immediately adjectival, each mode being What I shall attempt to say about the "grain" will, of course,
linked to a coded expression (harsh, austere, proud, virile, be only an apparently abstract approach, the impossible account-
solemn, majestic, warlike, educative, proud, ceremonious, ing of an individual enjoyment which I constantly experience
mourning, proper, dissolute, voluptuous); and among the Ro- when I listen to singing. In order to disengage this "grain"
mantics, from Schumann to Debussy, who substitute or add to from the acknowledged values of vocal music, I shall employ
the simple indication of movements (allegro, presto, andante) a double opposition: the theoretical one of the pheno-text and
certain poetic, emotive, increasingly refined predicates—given the geno-text (Kristeva's terms), and the paradigmatic one of
in the vernacular, so as to diminish the coded imprint and to two singers, one of whom I like very much (though he is no
develop the "free" character of the predication (sehr krdftig, longer to be heard) and the other very little (though he is
sehr pra'cis, spirituel et discret, etc.). heard more than anyone else): Panzera and Fischer-Dieskau
Are we doomed to the adjective? Are we faced with this (who will, of course, be no more than ciphers here: I am not
dilemma: the predicable or the ineffable? To know whether deifying the first and I have nothing against the second).
there are (verbal) means of talking about music without ad-
jectives, we would have to consider a little more closely all of Listen to a Russian bass (a church bass: for opera, the entire
music criticism, which, I believe, has never been done and voice shifts to dramatic expressivity: a voice in which the grain
which, even so, we have neither the intention nor the means signifies little): something is there, manifest and persistent
 The Grain of the Voice / 271
270 / ROLAND BARTHES
tion alien to communication, to representation (of feelings), to
(you hear only that), which is past (or previous to) the mean- expression; it is that culmination (or depth) of production
ing of the words, of their form (the litany), of the melisma where melody actually works on language—not what it says but
and even of the style of performance: something which is the voluptuous pleasure of its signifier-sounds, of its letters:
directly the singer's body, brought by one and the same move- explores how language works and identifies itself with that labor.
ment to your ear from the depths of the body's cavities, the Geno-song is, in a very simple word which must be taken quite
muscles, the membranes, the cartilage, and from the depths of seriously: the diction of language.
the Slavonic language, as if a single skin lined the performer's From the point of view of pheno-song, Fischer-Dieskau is
inner flesh and the music he sings. This voice is not personal: certainly an irreproachable artist; everything in the (semantic
it expresses nothing about the singer, about his soul; it is not and lyric) structure is respected; and yet nothing seduces, noth-
original (all Russian basses have this same voice, more or less), ing persuades us to enjoyment; this is an excessively expressive
and at the same time it is individual: it enables us to hear a art (the diction is dramatic, the caesuras, the checks and re-
body which, of course, has no public identity, no "personality," leases of breath intervene as in the upheavals of passion) and
but which is nonetheless a separate body; and above all this thereby it never transcends culture: here it is the soul that
voice directly conveys the symbolic, over and above the intel- accompanies the song, not the body: for the body to accompany
ligible, the expressive: here, flung before us all in a heap, is the the musical diction, not by an impulse of emotion but by a
Father, his phallic status. That is what the "grain" would be: "gesture-notice"*—that is what is difficult, especially since all
the materiality of the body speaking its mother tongue: per- musical pedagogy teaches not the culture of the "grain" of the
haps the letter; almost certainly what I have called signifying voice but the emotive modes of its emission: this is the myth
[signifiance]. of breath. How many singing teachers have we heard prophesy
It is here in song, then (pending the extension of the distinc- that the whole art of song was in the mastery, the proper man-
tion to all music), that we first discern the two texts of which agement of breathing! Now, the breath is the pneuma, the soul
Julia Kristeva writes. The pheno-song (if I may be permitted to swelling or breaking, and any exclusive art of the breath is
make this transposition) covers all the phenomena, all the likely to be a secretly mystical art (a mysticism reduced to the
features which derive from the structure of the sung language, demands of the long-playing record). The lung, a stupid organ
from the coded form of the melisma, the idiolect, the com- (the lights of catfood!), swells but does not become erect: it is
poser, the style of interpretation: in short, everything which, in the throat, site where the phonic metal hardens and assumes
in the performance, is at the service of communication, of its contour, it is in the facial mask that signifying breaks out,
representation, of expression: what is usually spoken of, what producing not the soul but enjoyment. In Fischer-Dieskau's
forms the tissue of cultural values (the substance of acknowl- performance, I seem to hear only the lungs, never the tongue,
edged tastes, of fashions, of critical discourse), what is directly the glottis, the teeth, the sinuses, the nose. Panzera's entire art,
articulated around the ideological alibis of a period (an artist's on the contrary, was in the letters, not in the bellows (a simple
"subjectivity," "expressivity," "dramaticism," "personality").
* "This is why the best way to read me is to accompany the reading
The geno-song is the volume of the speaking and singing voice, with certain appropriate body movements. Contra non-spoken writing,
the space in which the significations germinate "from within the contm non-written speech. Pro gesture-notice" (Philippe Sollers, Lois).
language and in its very materiality"; this is a signifying func-
 272 / R O L A N D B A R T H E S The Grain of the Voice / 273

technical feature: we did not hear him breathe, but only shape culture historically. Fischer-Dieskau reigns today almost ex-
the phrase). An extreme thought controlled the prosody of clusively over the long-playing song discography; he has re-
enunciation and the phonic economy of the French language; corded everything: if you like Schubert and you don't like
certain prejudices (generally resulting from oratorical and ec- Fischer-Dieskau, Schubert is inaccessible to you nowadays: an
clesiastical diction) were reversed. The consonants, which are example of that positive censorship (by repletion) which char-
too readily assumed to constitute the armature of our language acterizes mass culture without its ever being criticized for it;
(though it is not a Semitic language) and which we are always perhaps this is because Fischer-Dieskau's art, expressive, dra-
supposed to "articulate," to separate, to emphasize in order to matic, emotionally clear, conveyed by a voice without "grain,"
fulfill the clarity of meaning—these consonants Panzera fre- without signifying weight, corresponds perfectly to the re-
quently recommended skating over, restoring to them the ero- quirements of an average culture; this culture, defined by the
sion of a language which has lived, functioned, and worked for extension of listening and the disappearance of practice (no
a very long time, to make it into the simple springboard of the more amateur performers), is eager for art, for music, provided
admirable vowels: here was the "truth" of language, not its that such art and such music be clear, that they "translate" an
functionality (clarity, expressivity, communication); and the emotion and represent a signified (the poem's "meaning"):
range of the vowels received what was signifying (which is an art which vaccinates enjoyment (by reducing it to a known,
everything that can be voluptuous in meaning): the opposi- coded emotion) and reconciles the subject with what, in music,
tion of e and e (so necessary in conjugation); the virtually elec- can be said: with what is said of it, predicatively, by the Acad-
tronic purity, I should say, so taut, raised, exposed, tenuous was emy, by Criticism, by Opinion. Panzera does not belong to
its sound, of the most French vowel of all, the ii, which our this culture (he could not have done so, having sung before
language does not inherit from Latin; in the same way, Panzera the advent of the long-playing record; I doubt, moreover, that
held his r's beyond the singer's norms—without rejecting those his art, if he were singing nowadays, would be acknowledged
norms: his r was certainly rolled, as in any classical art of song, or even simply perceived); his reign—very widespread between
but such rolling had nothing peasant-like or Canadian in it; it the wars—was that of an exclusively bourgeois art (i.e., in no
was an artificial roll, the paradoxical state of a letter sound at way petit-bourgeois), concluding the fulfillment of its internal
once quite abstract (by its metallic brevity of the vibration) and development, separated from History—by a very familiar dis-
quite material (by its obvious implantation in the moving tortion; and it is perhaps, precisely and less paradoxically than
throat). Such phonetics (Am I alone in hearing it? Am I hear- it would seem, because this art was already marginal, mandarin,
ing voices in the voice? But is it not the truth of the voice to be that it could show traces of signifying [signifiance], could escape
hallucinated? Is not the entire space of the voice an infinite the tyranny of signification.
space? No doubt this was the meaning of Saussure's work on
anagrams)—such phonetics does not exhaust signifying (which The "grain" of the voice is not—or not only—its timbre; the
is inexhaustible); at least it imposes a limit on those efforts of signifying it affords cannot be better defined than by the fric-
expressive reduction made by a whole culture upon the poem tion between music and something else, which is the language
and its melody. (and not the message at all). The song must speak, or better
It should not be excessively difficult to date, to specify this still, must write, for what is produced on the level of geno-song
 274 / R O L A N D B A R T H E S The Grain of the Voice / 275

is ultimately writing. This sung writing of the language is, to ance of this death has to be dramatic: this is the triumph of
my sense, what the French melodie has occasionally attempted the pheno-text, the smothering of signifying under the signi-
to achieve. I know of course that the German lied, too, has fied: soul. Melisande, on the contrary, dies only prosodically;
been intimately linked with the German language by the in- two extremes are linked, braided: the perfect intelligibility of
termediary of the romantic poem; I know that Schumann's the denotation, and the pure prosodic contour of the enuncia-
poetic culture was vast and that this same Schumann once said tion: between the two a beneficent void, which constituted
of Schubert that if he had lived to be old he would have set all the repletion of Boris: pathos, i.e., according to Aristotle (why
of German literature to music; but all the same I think that the not?), passion as men speak it, imagine it, the accepted notion
historical meaning of the lied must be sought in its music (if of death, endoxal death. Melisande dies without any noise; let
only by reason of its folk origins). On the contrary, the his- us understand this expression in its cybernetic sense: nothing
torical meaning of the French melodie is a certain culture of comes to disturb the signifier, and hence nothing compels
the French language. We know that the romantic poetry of redundancy; there is production of a music-language whose
our country is more oratorical than textual; but what our poetry function is to prevent the singer from being expressive. As for
has not been able to do in and of itself, the melodie has some- the Russian bass, what is symbolic (death) is immediately cast
times accomplished in collaboration with it; it has worked on (without mediation) before us (this in order to forestall the
the language through the poem. This work (in the specificity accepted notion according to which what is not expressive has
here granted to it) is not apparent in the general run of melod- to be cold, intellectual; Melisande's death is "moving"; this
ic production, overindulgent as it is of minor poets, of the model means that it moves something in the chain of the signifier).
of the petit-bourgeois ballad, of salon practice; but it is in- The French melodie has disappeared (one might even say that
controvertible in several works: anthologically (let us say: it sank like a stone) for a good many reasons, or at least this
virtually at random) in certain melodies by Faure and Duparc, disappearance has assumed a good many aspects; it doubtless
massively in the late (prosodic) Faure and in Debussy's vocal succumbed under the image of its salon origin, which is some-
oeuvre (even if Pelleas is often sung badly: dramatically). What what the parody of its class origin; classical music of mass
is involved in these works is much more than a musical style, culture (radio, records) has left it behind, preferring either the
it is a practical reflection (so to speak) on the language; there more emotive orchestra (Mahler's triumph) or instruments
is a gradual assumption of the language to the poem, of the less bourgeois than the piano (harpsichord, trumpet). But,
poem to the melodie, and of the melodie to its performance. above all, this death accompanies a much greater historical
This means that the French melodie derives very little from phenomenon, one which has little connection with the history
the history of music and a great deal from the theory of the of music or with that of musical taste: the French are abandon-
text. The signifier must, here again, be redistributed. ing their language, not of course as a normative set of noble
Let us compare two sung deaths—both very famous—that of values (clarity, elegance, correctness)—or at least we scarcely
Boris, that of Melisande. Whatever Musorgsky's intentions, concern ourselves much about that, for those are institutional
Boris's death is expressive, or one might even say, hysterical; it values—but as a space of pleasure, of enjoyment, a site where
is overloaded with affective, historical content; every perform- the language works upon itself for nothing, i.e., in perversion
 2j6 / R O L A N D B A R T H E S The Grain of the Voice / 2jy

(let us recall here the singularity—the solitude—of Philippe quite illusory, moreover), almost all of which belong to the
Soller's recent text Lois, which re-presents the prosodic and pheno-song (I shall not go into ecstasy over the "rigor," the
metrical work of the language). "brilliance," the "warmth," the "respect for the score," etc.),
but according to the image of the body (the figure) which is
The "grain" is the body in the singing voice, in the writing given me: I hear without a doubt—the certitude here of the
hand, in the performing limb. If I perceive the "grain" of this body, of the body's enjoyment—that Wanda Landowska's
music and if I attribute to this "grain" a theoretical value harpsichord comes from her inner body, and not from the
(this is the assumption of the text in the work), I cannot help minor digital knitting of so many harpsichordists (to the point
making a new scheme of evaluation for myself, individual no where I hear a different instrument); and with regard to piano
doubt, since I am determined to listen to my relation to the music, I know immediately which part of the body it is that
body of someone who is singing or playing and since that rela- plays: if it is the arms, as all too often it is, muscular as a
tion is an erotic one, but not at all "subjective" (it is not the dancer's calf, or the talons (regardless of the wrist flourishes),
psychological "subject" in me who listens; the enjoyment that or if on the contrary it is the only erotic part of a pianist's body:
subject seeks is not going to reinforce him—to express him— the pads of the fingers, whose "grain" I hear so rarely (need I
but on the contrary will destroy him). This evaluation will be remark that there seems to be, nowadays, under the pressure
made outside of the law: it will baffle the law of culture but of long-playing records and their mass sales, a flattening out of
also the law of anti-culture; it will develop beyond the subject technique, which is paradoxical: all playing is flattened out into
all the value which is hidden behind "I like" or "I don't like." perfection: there is nothing left but pheno-text).
Singers, particularly, will be ranked in two categories which we All of this has been said about "classical" music (in the
might call prostitutive, since it is a matter of my choosing what broadest possible sense); but it goes without saying that the
does not choose me: hence, I shall freely exalt some little- mere consideration of the musical "grain" could lead to another
known, secondary, forgotten, even dead artist, and I shall turn history of music than the one we know (which is purely pheno-
away from some consecrated star (let us not furnish examples, textual): if we were to succeed in refining a certain "aes-
they would doubtless have only a biographical value), and I thetic" of musical enjoyment, we should doubtless attach less
shall shift my choice to every genre of vocal music, including importance to the tremendous break in tonality which modern-
popular music, where I shall have no difficulty recognizing the ity had produced.
distinction between pheno-song and geno-song (certain artists
in this genre have a "grain" which others, however celebrated,
1972
do not). Further, aside from the voice, in instrumental music,
the "grain" or its lack persists; for if there is no longer a lan-
guage here to afford signifying in its extreme form and scope,
there is at least the artist's body which once again compels me
to an evaluation: I will not judge a performance according to
the rules of interpretation, the constraints of style (which are