Académique Documents
Professionnel Documents
Culture Documents
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Yale University Press is collaborating with JSTOR to digitize, preserve and extend access to Yale French
Studies.
http://www.jstor.org
Barbara Johnson
Poetry and PerformativeLanguage
Barbara Johnson
words, when, accordingto this text, does it make sense that there
be a poem?
On its most obvious level, La Declaration foraineis the story
of an improvisedside-show composed of two parts: a motionless
woman and a spoken poem. In the context,the relation between
the two seems deceptivelytransparent: the sonnet ("La chevelure
vol d'une flamme...") is simply,as R. G. Cohn describes it, "a
celebration of a woman whose looks, featuringmagnificenthair,
need no outer adornment."1 The poet's act would thus seem to
that "all thingsin life
bear out Remy de Gourmont'saffirmation
having been said thousands and thousands of times,the poet can
no longerdo anythingbut point to them,accompanyinghis gesture
with a few murmuredwords."2 In its simultaneousact of naming
and exhibiting,the poem can thus be said to relate to the lady as
a sign to its referent.
But if that is the case, how does this poem fitin with the rest
of Mallarme's poetics of "suggestion,"which he explicitlyopposes
to literal denomination?I If one recalls Mallarme's repeated inon the "vibrasistenceon poetry'sabolitionof simplereferentiality,
4
tory near-disappearance" of the real object "on which the pages
would have trouble closing,"' one begins to suspect both that the
is inadequate,
traditionalreading of Mallarme's non-referentiality
and that the lady's hair is only the apparentsubject of the sonnet,
or "surface"meaningwhichboth hides and reveals
the "indifferent"
somethingto which it remains"exterior."6 Mallarme's own highly
between the obambiguous statementof the non-correspondence
vious and the true in his own work is probablyresponsiblefor the
universalcriticaltendencyto give the hair a symbolicmeaning,to
findthe "pure notion" or "idea"-Poetry, ideal Beauty,naked truth,
1 RobertG. Cohn, Toward the Poems of Mallarme(Berkeley: University
of CaliforniaPress, 1965), p. 147.
2 Remy de Gourmont,Promenadeslittgraires
(IVeserie, 1912), p. 8. All
translationsfromthe French are my own.
3 Cf. Mallarm6, OEuvresCompletes (Paris, Bibliothequede la Pl6iade,
1945), pp. 366, 645, 869.
4 Ibid., p. 38.
5 Ibid., p. 366.
6 Ibid., p. 382.
141
142
Barbara Johnson
The story of the recitationof an occasional poem thus concludes with a discussion of what constitutesa poem's occasion:
the two ex-performers
are interestednot in what the poem means,
but in how it means, and in how it managed to come into being
at all. Two conditions,whose significancewe will discuss later,appear necessaryforthe poem to occur: audience and violence.Without them, the poet would "perhaps, who knows? not have introduced the pretext of formulating"his poem into the silent,
isolated togethernessof the rockingcoach. In fact,the prose poem,
which thus ends by discussing the necessary conditions for the
productionof speech,beginsby triplyinsistingon a state of absence
of speech:
Le Silence! il est certainqu'h mon c&t6,ainsi que songes,6tenduedans
un bercementde promenade sous les roues assoupissantl'interjectionde
fleurs, toute femme, et j'en sais une qui voit clair ici, m'exempte de l'effort
a proferer un vocable: la complimenter haut de quelque interrogatrice toi-
143
Yale FrenchStudies
This list
bation),and even adjectives(interrogatrice,
appreciative).
resemblesnothingso much as the concludingchapterof J. L.
Austin's How to Do Things with Words, in which an attemptis
144
BarbaraJohnson
inquires,"When we issue any utterancewhatsoever,
are we not
12 In his attempts
'doingsomething'?"
to finda formulainclusive
of all speech acts in whichsayingis doing,Austinpasses from
considerations
of grammatical
formand transformational
rulesto
considerations
of semanticcontentand interpersonal
effects.
In the
course of the inquiry,the originalbinaryoppositionbetween
performative
and constativelanguageinevitably
breaksdown.The
impossibility
ofdefining
thelinguistic
specificity
oftheperformative
utterance(forwhichwe will tryto accountlateron) leads Austin
to drawup a new set analytictermsfocusingnot on the intrinsic
characteristics
of an utterancebut on its actual functionin an
interlocutionary
situation.Abandoning
the performative/constative
Austinproposesto analyzeany utterance
dichotomy,
accordingto
three"dimensions":1) thelocutionary (sound,senseand reference),
a) theillocutionary(intentional
and conventional
force),and 3) the
perlocutionary (actual effect).
Sincethesenotions,thoughnot withouttheirusefulness,
are at
least as problematic
as the notionof the performative,
subsequent
thinkers
have preferred
to returnto the searchfora set of stable
linguisticcriteriafor the isolationof the performative
itself.By
choosingthesecriteriain such a wayas to eliminateall but what
Austinhimself
calls "explicitperformatives",
thistaskbecomesrelareverbsin thefirst
ativelysimple:explicitperformatives
(orimpersonal third)personsingularpresentindicative
activewhichpossess
"an asymmetryof a systematic
kind[withrespectto] otherpersons
and tenses of the verysame word."13 That is, to use Austin's exam-
whereas
ple,"I bet" is theactualperformanceofthe act ofbetting,
The performative
is
"he bets"is onlya reportof an act of betting.
ofuttering
is "at themoment
iftheactionperformed
onlyoperative
14 The performative,
beingdone by the personuttering."
then,acts
in thatit takeson meaningonlyby referring
to the
like a "shifter"
instanceof its utterance.The FrenchlinguistEmile Benveniste,
12 Ibid., p.
13 Ibid., p.
14 Ibid., p.
92.
63.
60.
145
Yale FrenchStudies
semanticdimensionto the definition,
adding a self-referential
eliminatesany remaining
whenhe asserts:
effectively
uncertainty
"An utteranceis performative
insofaras it names the act perthe act
formed....The utterance
is the act; the uttererperforms
by namingit." 15
("proposa,""consentit,"
"accepta").Thatis,thespeechact to which
theyreferis notbeingperformed
but onlynamedor reported.
The
name of the de-activated
speechact therefore
functionslike any
othernoun,evento thepointofserving
as a metaphor
forsomething
totallyunrelatedto a literalspeechact ("l'interjectionde fleurs,"
15 Benveniste,Problemes de linguistique generate (Paris: Gallimard,
1966), p. 274.
146
BarbaraJohnson
"commeune vociferation").
Thus, if a performative
utteranceis
originallya self-referential
speech act, its productionis simultaneouslythe productionof a new referent
into the world.This,
however,is tantamount
to a radicaltransformation
of the notion
of a referent,
since,insteadof pointingto an externalobject,languagewouldthenreferonlyto its ownreferring
to itselfin theact
of referring,
and the signifying
chain would end in an infinitely
self-duplicating
loop. A variantof thisdifficulty
has, in fact,been
pointedout by P. Larreya,who,in attempting
to fita performative
utteranceinto a Chomskyan
tree diagram,findsthat "to develop
the
thetreeit wouldbe necessary
to repeatthesymbol[designating
an infinitenumberof times."16 The performative
performative]
utterance
is thusthe miseen abymeof reference
itself.
We havenow arrivedat a predicament
similarto thatdescribed
17 but we
by RichardKleinin his studyof metaphors
of metaphor,
are stilla longwayfromshowingwhatthepoemhas to say about
the relationbetweenthis predicament
of
and the characteristics
languagein general.In pursuitof this question,let us examine
of the persome furtherimplicationsof the self-referentiality
utterance.If the performative
formative
refersonly to itself,it
wouldseem thatit does not referto any exterioror priororigin.
In actualanalysis,however,we see thatthisis neverconsidered
to
be the case. For althoughthesenseand the reference
of thespeech
the presence
act are its own utterance,
thatveryfactpresupposes
of theutterer,
whothenbecomesthenecessary
originofthespeech
act in question.Somesignofthespeaker'spresenceto hisutterance
utteranceis to be
if the performative
is consideredindispensable
But in thisprosepoem,it is precisely
whatAustincalls "felicitous."
the intentional
continuity
betweenthe speakerand the utterance
whichis beingquestionedbythepoetandhislady,forthe"reverie"
whichhas been proclaimed
to the crowd"s'ignoreet se lancenue
de peur,"just as the call to the crowdto enterthe boothin the
16 Paul Larreya,"Enoncds performatifs,
cause, et r6ference,"in Degres,
jere annie, No 4, Oct. 1973, p. m23.
17 Richard Klein, "StraightLines and Arabesques: Metaphorsof Metaphor," in Yale French Studies, 45 (Language as Action), 1970.
147
Yale FrenchStudies
firstplace had been "obscurpour moi-memed'abord."18 Indeed,
if whatthe poet has spewedforth"force,sous le coup de poing
it is
brutal'a l'estomac"can in any way be called self-expression,
senseof the word.
so onlyin theetymological
toothpaste-tube-like
The poemis notgenerated
naturally
by thepoet'ssubjectiveintenfromthe poet's mouthuntimely
tionality:it is, on the contrary,
withMallarme'smuchripped.This,of course,is totallyconsistent
ofthepoeticsubject: "L'aeuvrepureimplique
discussedelimination
la disparition6locutoiredu poete, qui cede l'initiative aux
mots..."19 Indeed, the active productionof this discontinuity
the perbetweenthe speakerand his words,farfromeliminating
formative
dimensionin Mallarme'spoetry,may itselfconstitute
thatpoetry'strulyrevolutionary
performativity.
However,ifwe returnnowto thewayin whichthiselimination
is actuallyevoked at the end of La Declaration
of subjectivity
oftherelationofspeaker
we findthateventhisformulation
foraine,
to speech is oversimplified.
For the assertionof the non-intentionality
of the poemis itselfso tortuously
non-committal
thatby
the time it ends in an unequivocal"c'est vrai," it has already
practicallyqualifieditselfout of existence.While namingthe
impatience
ofthecrowdas theexplicit"cause"onlyofthefigurative
"punchin thestomach"whichmakesthepoem"squirtout" of the
poet,theladyneithertotallyexcludesthepossibility
of thepoem's
havingoccurredin the carriage(into whichthe poet would only
perhaps,who knows?not have introduced
it), nor does she articulate in any way the relationbetweenpunchand squirt,which
cannotevenbe said to meeton thesamerhetorical
level.
Turningto the circumstances
surrounding
the utteranceof the
one true performative
expression"je le gage,"we finda similar
18 As Ursula Franklinpoints out (The Prose Poems of Stephane
Mallarme: An Exegesis, Michigan State University,Ph. D. Dissertation,1971,
reproduced by UniversityMicrofilms),the period after "d'abord" in the
Pleiade editionis a typographical
error: it is the speech itselfwhichis being
modifiedby the expression"invariableet obscur."
19 iEuvrescompletes, p. 366: "The pure
(poetic) work implies the
elocutionary disappearance of the poet, who leaves the initiative to
words. . ."
148
Barbara Johnson
problematizationof the nature of the act performed.For if, accordingto Austin,a bet can only be said to occur if it is accepted
by a taker, the "peut-etre!" with which this taker "accepts" the
poet's bet effectivelysuspends its application and thus its ability
to functionas a true act. Moreover,what is or is not being wagered
here seems itselfinternallyinconsequent,since the "irrefutability"
of the poet's spiel is dependentnot on the clear univocalityof its
meaning,but, on the contrary,on the uncontrollablemultiplicity
of its repercussions.
Thus, while "c'est vrai" and "je le gage" explicitlymark the
places of the constative and the performativerespectively,what
happens in between is that what is stated is the problematization
of the conditions of performance,while what is wagered is the
problematizationof the possibilityof statement.
Austin's theory,of course, contains no provisionfor this type
of ambiguity.Its eliminationis, in fact, one of the main motives
behind the explicitationof a performativeexpression,since "the
explicit performativerules out equivocation."20 But behind the
question of ambiguity,somethingmuch more unsettlingis at stake,
forit is not onlyequivocationwhichis ruled out by Austin's discussion of performative
utterances: it is nothingless than poetryitself.
149
23
150
Barbara Johnson
thus automaticallyfictionalizesits uttererwhen it makes him the
mouthpieceof a conventionalizedauthority.Where else, for example, but at a Party Conventioncould a presidentialcandidate be
nominated? Behind the fictionof the subject stands the fictionof
Society,24 for if one states that Society began with a prohibition
(of incest) or a (social) contract,one is simplystatingthat the origin
of the authoritybehind a performative
utteranceis derivedfroma
previous performativeutterancewhose ultimate originis undeterminable. By using these tu quoque arguments,it is, of course, not
our intentionto nullifyall differencesbetween a poem and, say,
a verdict,but only to problematizethe assumptionson which such
distinctionsare based. If people are put to death by a verdictand
not by a poem, it is not because the Law is not a fiction.
The non-seriousnessof a performativeutterance "said by an
actor on the stage" results,then,not fromhis fictionalstatus but
fromhis duality,fromthe spectator'sconsciousnessthat although
the characterin the play is swearingto avenge his dead father's
ghost,the actor's own performative
commitments
lie elsewhere.But
the performative
utteranceitselfis here just as "serious" withinthe
context of its surroundingfictionas it would be in the contextof
the fictionwe call real life. Indeed, the question of "seriousness"
attends the act of interpretationof any performativeutterance
whatever.Rhetorical imperatives,for example, are far frombeing
restrictedto poetry: a large proportionof our ordinaryconversational devices consists of such expressionsas "Go jump in a lake,"
"Go fly a kite," and other more frequentbut less mentionable
retorts.The question of seriousness,far frommarkingthe borders
is foundto inhabitthe verycore if its territory.
of the performative,
This is, in fact, one of the main factorsbehind Austin's recourse
to the notionof illocutionaryforce.And thisquestion,as it happens,
returnwould inevitablybe reversedby language- in this case by the suspension of an act by the word "no," utterednot by the victimbut by the
perpetratorof the act.
24 Cf. Mallarm6'sdescriptionof the word "Soci&V": "La Soci6t6,terme
le plus creux, h6ritagedes philosophes,a ceci, du moins, de propice et
d'ais6 que rien n'existant, 'a peu pres, dans les faits,pareil l'injonction
qu'6veille son concepte auguste, en discourir,6gale ne traiteraucun sujet
ou se taire par d6lassement."(in "Sauvegarde,"(Euvres completes,p. 419).
151
The hair
from
fight of a flame to
the far
theft
at
vvd
this alive
aie
should continue
continues
that is continuous
152
nd
cude
lu
Barbara Johnson
Here, it is the "ceil veridique ou rieur,"roughlyequivalent to the
''naive or ironic reader," which raises the question of seriousness.
in termsof an alternative
By namingthe problemof interpretation
between seriousness and irony,the sonnet places itself between
two incompatiblereadings of its own illocutionaryforce. Readers
of La Declaration foraineare indeed oftensensitiveto the mocking
way in which the poet seems to treathis own creation: in her very
helpfuldiscussionof this prose poem, Ursula Franklin,forexample,
uses the word "irony" and its derivativesno less than fourteen
times. But behind the question of illocutionaryforcelies the queswhich,as we have alreadyseen, is here being
tion of intentionality,
subvertedby the involuntary,blind relationshipbetween the poet
and his poem. We would thereforeexpect that the sonnet itself
would somehowescape the simpledichotomythat it evokes between
seriousness and irony,as, indeed, the poet says it does when he
speaks of its "comprehensionmultiple."Let us thereforeexamine
the text of the poem in order to follow the precise functioning
of this interpretativemultiplicity.
This poem has been "read" many times.26There seems to be
little doubt that it is "about" the woman standingbehind it, and in
particular,about her hair. But if one attemptsto make explicitnot
the referenceitselfbut the sense of the reference-whatthe poem
is saying about the woman-one findsthat the actual affirmations
made by the poem are very difficultto pin down. In attempting
strategies-the isolato pursue even the simplestof interpretative
tion of all the verbs in the presenttense,forexample-one stumbles
over the word "continue," which may be not a verb but an
adjective,and even if it is a verb may be eithertransitiveor intransitive. But a tentativegrammaticalskeleton mightrun something
like this:
26 Cf., in addition to the excellentlist givenby Ursula Franklin(op. cit.
pp. 230-231): Charles Mauron, Mallarmgl'obscur (Paris: Corti, 1968) and
Austin Gill, Mallarme'sPoem "La Chevelurevol d'une flamme..." (University of Glasgow, 1971).
153
Barbara Johnson
poem? Isn't the lady's "exploit"still being presentedas a dazzlingly
self-evidentact in its own right? The poet's partingwords to the
crowd, indeed, appear to be sayingjust that:
La personne qui a eu 1'honneurde se soumettre'a votre jugement,ne
requiertpour vous communiquerle sens de son charme,un costumeou aucun accessoire usuel de theatre.Ce naturels'accommodede I'allusion parfaite que fournitla toilettetoujours a l'un des motifsprimordiauxde la
27
femme,et suffit.
155
completes p. 367).
156
Barbara Johnson
comes "true" only throughits radical incompatibilitywith itself.
As we have seen, this "suspension" of meaning may occur
through the simultaneous presence of contradictoryaffirmations,
But if, as in the case of the word "diffame,"the play of contradictions lies in the very separation ("dehiscence") of a word from
itself,this is a highlyunsettlingfactor.The very diachronywhich
has moved "diffamer"from the constative "divulge" to the performative"malign" is at work in any utterancewhatsoever: quite
apart fromthe question of "seriousness,"for example, the illocutionaryforceof an utteranceis subject to the same kind of temporal
fading and conventionalizingthat produces "dead" metaphorsand
cliches. Benveniste'sattemptto exclude "simple formulas"like "je
m'excuse" and "bonjour" fromconsiderationas "live" performatives
is doomed by the very nature of "living" language itself.
That the logic of language renders some kind of discontinuity
between speaker and speech absolutely inescapable is in fact demonstratedpreciselyby Austin's very attemptto eliminateit. For
the very word he uses to name "mere doing," the very name he
gives to that fromwhich he excludes theatricality,is none other
than the word whichmost commonlynames theatricality:the word
"perform."As if this were not ironic enough,the exact same split
can be foundin Austin's otherfavoriteword: the word "act." How
is it that a word which expresses most simplythe mere doing of
an act necessarilyleads us to the question of... acting? How is it
possible to discuss the question of authenticitywhen that question
alreadysubvertsthe verytermswe use to discuss it? Is it inevitable
that the same split that divides the referentfromitselfthe moment
language comes near it should divide language from itself in the
very same way? And can language actually referto anythingother
than that very split? If Austin's unstated question was "What are
egaux par valeur et 'a les grouper; ils riment: pour sceau exterieur,leur
157
158