The Frame of Reference, Poe, Lacan Derrida - Barbara Johnson

The Frame of Reference: Poe, Lacan, Derrida
Author(s): Barbara Johnson

Source: Yale French Studies, No. 55/56, Literature and Psychoanalysis. The Question of Reading:
Otherwise (1977), pp. 457-505
Published by: Yale University Press
Stable URL: http://www.jstor.org/stable/2930445
Accessed: 05-11-2015 01:26 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
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
This content downloaded from 143.167.2.135 on Thu, 05 Nov 2015 01:26:03 UTC
All use subject to JSTOR Terms and Conditions
Barbara Johnson
The Frame of Reference: Poe, Lacan, Derrida
1. The Purloined Preface
A literarytext which both analyzes itselfand shows that it ac-

tually has neithera self nor any neutral metalanguagewith which
to do the analyzing,calls out irresistiblyfor analysis. And when
thatcall is answeredby two eminentFrenchthinkerswhose readings
emit an equally paradoxical call-to-analysisof theirown, the result-
ing triptych,in the context of the question of the act-of-reading
(-literature),places its would-be reader in a vertiginouslyinsecure
position.
The three texts in question are Edgar A. Poe's short storyThe
Purloined Letter,' Jacques Lacan's Seminar on The Purloined
Letter,2 and Jacques Derrida's reading of Lacan's reading of Poe,
The Purveyorof Truth[Le Facteur de la Verite].I In all threetexts,
it is the act of analysis which seems to occupy the center of the
discursivestage,and the act of analysis of the act of analysis which
in some way disruptsthat centrality.In the resultingasymmetrical,
abyssal structure,no analysis-including this one-can intervene
withouttransforming and repeatingother elementsin the sequence,
which is thus not a stable sequence, but which neverthelesspro-
duces certainregulareffects.It is the functioningof this regularity,
and the structureof these effects,which will provide the basis for
the present study.
I In Great Tales and Poems of Edgar Allan Poe, The Pocket Library,
New York, 1951, hereafterdesignatedas "Poe."
2 In tcrits (Paris: Seuil, 1966). Quotations in English are taken,
unless otherwise indicated, from the partial translationin Yale French
Studies 48, French Freud, 1973, hereafterdesignatedas "SPL."
3 This article was published in French in Poetique 21 (1975) and,
somewhat reduced, in Yale French Studies 52, Graphesis, 1975. Unless
otherwise indicated, references are to the English version, hereafter
designatedas "PT."
457
Yale FrenchStudies
The subversionof any possibilityof a positionof analytical

mastery occursin manyways.Here,theveryfactthatwe are deal-
ingwiththreetextsis in no waycertain.Poe's storynot onlyfits
intoa triptych of its own,but is riddledwitha constant, peculiar
kindof intertextuality (theepigraphfromSenecawhichis notfrom
Seneca, the lines fromCrebillon'sAtree whichserve as Dupin's
signature,
etc.).Lacan's textnot onlypresentsitselfbackwards(its
introductionfollowing its conclusion),
but it neverfinishes
present-
ing itself ("Ouverture de ce recueil," "Presentationde la suite,"
"Presentation" to thePointsedition).And Derrida'stextis notonly
precededby severalyearsofannunciatory marginaliaand footnotes,
but it is itselfstructured preciselyby its own deferment, its dif-
ferance(cf. the repetition of such expressionsas "Mais nous n'en
sommespas encorela," "Butwe aregetting aheadofourselves," etc.).
In addition,all of these textsare characterized by an unusually
highdegreeof apparentdigressiveness, to the pointof makingthe
readerwonderwhether thereis reallyanytruesubjectmatterthere
at all. It is as thoughanyattempt to followthepathofthepurloined
letteris automatically purloinedfromitself.Whichis, as we shall
see, just whatthe letterhas alwaysalreadybeen saying.
Any attemptto do "justice" to threesuch complextexts is
obviouslyout of thequestion.But it is precisely thenatureof such
"justice"thatis the questionin each of thesereadingsof the act
of analysis.The factthat the debateproliferates arounda crime
story-a robbery and itsundoing-canhardlybe an accident.Some-
where-in each of these texts,the economyof justicecannotbe
avoided.For in spiteof the absenceof mastery, thereis no lack of
effectsof power.
As the readergoes on withthisseriesof prefatory remarks,he
may begin to see how contagiousthe deferment of the subject
of the purloinedlettercan be. But the problemof how to present
thesethreetextsis all the moreredoubtablesince each of them
bothpresentsitselfand the others,and clearlyshowsthe fallacies
inherent in anytypeof "presentation" of a text.The factthatsuch
fallaciesare not onlyinevitablebut also constitutive
of anyact of
458
BarbaraJohnson
reading-alsodemonstrated by each of thetexts-is smallcomfort,

since the resultinginjustices,howeverunavoidablein general,al-
ways appear corrigiblein detail. Which is why the sequence
continues.
The questionofhowto presentto thereadera texttoo extensive
to quote in its entirety
has in factlongbeen one of theunderlying
problemsof literarycriticism.Since a shorterversionof the text
mustsomehowbe produced,two solutionsconstantly recur: para-
phraseand quotation.Althoughthese tacticsare seldomif ever
used in isolation,the specificconfigurationof theircombinations
and permutations determines to a large extentthe "plot" of the
criticalnarrativeto whichtheygiverise.The firstact of our own
narrative,then,will consistof an analysisof the strategiceffects
of the use of paraphrasevs. quotationin each of the threetexts
in question.
2. Round Robbin'
Round robin: 1) A tournamentin whicheach
contestant is matched against every other
contestant.2) A petitionor proteston which
the signaturesare arrangedin the formof a
circlein orderto conceal the orderof signing.
3) A lettersent among membersof a group,
oftenwith commentsadded by each person
in turn. 4) An extended sequence.
-The AmericanHeritage Dictionary
In 1845,EdgarA. Poe publishedthethirdof his threedetective

stories,"The PurloinedLetter,"in a collectivevolumeentitled-
ironicallyenough,considering all the robberiesin the story-The
Gift:A Christmas, New Year, and BirthdayPresent."The Pur-
loinedLetter"is a first-person narrationof two scenes in which
dialoguesoccuramongthe narrator, his friendC. AugusteDupin,
and, initially,the Prefectof the Parisianpolice. The two scenes
are separatedby an indicationof the passageof a month'stime.
In each of the two dialogues,reportedto us verbatim by the nar-
rator,one of the othertwo characters tellsthe storyof a robbery:
459
Yale FrenchStudies
in thefirstscene,it is thePrefectofPolicewhorepeatstheQueen's
eyewitness accountof the Minister'stheftof a letteraddressedto
her; in the secondscene,it is Dupin who narrateshis own theft
of the same letterfromthe Minister,who had meanwhileread-
dressedit to himself.In a paragraphplaced betweenthese two
"crime"stories,the narratorhimselfnarratesa wordlessscene in
whichthe letterchangeshandsagainbeforehis eyes,passingfrom
Dupin-not withoutthelatter'shavingaddressednottheletterbut
a checkto himself-tothe Prefect(who will pocketthe remainder
of the reward)and thence,presumably, back to the Queen.
By thusappearingto repeatto us faithfully everywordin both
dialogues,the narrator wouldseemto have resortedexclusively to
directquotationin presentinghis story.Even when paraphrase
couldhavebeenexpected-inthedescription oftheexactprocedures
employedby the police in searchingunsuccessfully forthe letter,
forexample,-weare sparednoneof the details.Thus it is all the
moresurprising to findthatthereis one littlepointat whichdirect
quotationof the Prefect'swords gives way to paraphrase.This
point,howeverbrief,is of no small importance, as we shall see.
It occursin the concluding paragraphof the firstscene:
"I have no betteradvice to give you," said Dupin. "You have, of

course,an accuratedescriptionof the letter?"
"Oh, yes!" -And here the Prefect,producinga memorandum-book,
proceededto read aloud a minuteaccountof the internal,and especially
appearanceof the missingdocument.Soon afterfinishing
of the external,
the perusalof this description, moreentirely
he took his departure, de-
pressedin spiritsthan I had ever knownthe good gentlemanbefore.
(Poe, pp. 206-207.)
Whatis paraphrasedis thusthe description of the letterthe story

is about. And, whereasit is generallysupposedthatthe function
of paraphraseis to stripoffthe formof a speechin orderto give
us onlyits contents,herethe use of paraphrasedoes the veryop-
posite: it withholdsthe contentsof the Prefect'sremarks, giving
us onlytheirform.And what is swallowedup in this ellipsisis
nothingless than the contents of the letter itself. The fact that
460
BarbaraJohnson
theletter'smessageis neverrevealed,whichwillserveas thebasis

forLacan's readingof the story,is thus negatively made explicit
by the functioning of Poe's textitself,throughwhatDerridamight
have calleda repression of thewritten word(a suppression of what
is writtenin the memorandum-book-and in the letter).And the
questionof the strategicuse of paraphraseversusquotationbegins
to invadenotonlythecriticalnarrative, buttheliterarytextas well.
Lacan's presentationof Poe's text involvesthe paraphrase, or
plot summary, of the two theftsas theyare told to the narrator
by thePrefectand byDupin.SinceDerrida,in hiscritiqueofLacan,
choosesto quote Lacan's paraphrase, we can combineall thetactics
involvedby, in our turn,quoting Derrida'squotationof Lacan's
paraphraseof Poe's quoted narrations. 4
There are two scenes, the firstof which we shall straightwaydesignate

the primal scene, and by no means inadvertently, since the second may
be considered its repetitionin the very sense we are consideringtoday.
The primal scene is thus performed,we are told [by neitherPoe, nor
the scriptor,nor the narrator,but by G, the Prefect of Police who is
mis en scene by all those involved in the dialogues- J. D. 5] in the royal
boudoir, so that we suspect that the person of the highest rank, called
the "exalted personage," who is alone there when she receives a letter,
is the Queen. This feelingis confirmedby the embarrassmentinto which
she is plunged by the entryof the other exalted personage,of whom we
have already been told [again by G - J. D.] prior to this account that
the knowledge he might have of the letter in question would jeopardize
for the lady nothingless than her honor and safety.Any doubt that he
is in fact the King is promptlydissipated in the course of the scene
which begins with the entry of the Minister D.... At that moment,in
fact, the Queen can do no better than to play on the King's inattentive-
ness by leaving the letter on the table "face down, address uppermost."
It does not, however,escape the Minister'slynx eye, nor does he fail to
notice the Queen's distress and thus to fathom her secret. From then
on everythingtranspireslike clockwork. After dealing in his customary
mannerwith the business of the day, the Ministerdraws fromhis pocket
a lettersimilarin appearance to the one in his view, and, havingpretended
4 Such a concatenation could jokingly be called, after the nursery

rhyme,"This is the text that Jacques built." But in fact, it is precisely
this kind of sequence or chain that is in question here.
5 We will speak about this bracketed signature later; for the time
being it stands as a sign that Derrida's signaturehas indeed been added
to our round robin.
461
Yale FrenchStudies
to read it, places it next to the other. A bit more conversationto amuse
the royal company,whereupon,withoutflinchingonce, he seizes the em-
barrasingletter,making off with it, as the Queen, on whom none of his
maneuverhas been lost, remainsunable to intervenefor fear of attracting
the attentionof her royal spouse, close at her side at that very moment.
Everythingmight then have transpired unseen by a hypothetical
spectator of an operation in which nobody falters,and whose quotient
is that the Ministerhas filchedfrom the Queen her letter and that-an
even more importantresultthan the first-the Queen knows that he now
has it, and by no means innocently.
A remainderthat no analyst will neglect, trained as he is to retain
whateveris significant,withoutalways knowingwhat to do with it: the
letter,abandoned by the Minister,and which the Queen's hand is now
free to roll into a ball.
Second scene: in the Minister's office.It is in his hotel, and we
know-from the account the Prefect of Police has given Dupin, whose
specific genius for solving enigmas Poe introduces here for the second
time-that the police, returningthere as soon as the Minister'shabitual,
nightlyabsences allow them to, have searched the hotel and its sur-
roundingsfromtop to bottom for the last eighteenmonths.In vain,-al-
though everyonecan deduce from the situation that the Minister keeps
the letter within reach.
Dupin calls on the Minister. The latter receives him with studied
nonchalance, affectingin his conversation romantic ennui. Meanwhile
Dupin, whom this pretencedoes not deceive, his eyes protectedby green
glasses,proceeds to inspectthe premises.When his glance catches a rather
crumbled piece of paper-apparently thrust carelessly in a division of
an ugly pasteboard card-rack,hanging gaudily from the middle of the
mantelpiece-he already knows that he's found what he's looking for.
His convictionis reinforcedby the very details which seem to contradict
the descriptionhe has of the stolen letter, with the exception of the
format,which remains the same.
Whereuponhe has but to withdraw,after "forgetting"his snuff-box
on the table, in order to returnthe followingday to reclaim it -armed
with a facsimileof the letter in its present state. As an incident in the
street,preparedfor the propermoment,draws the Ministerto the window,
Dupin in turn seizes the opportunityto seize the letterwhile substituting
the imitation,and has only to maintainthe appearances of a normal exit.
Here as well all has transpired,if not withoutnoise, at least without
all commotion.The quotient of the operation is that the Minister no
longer has the letter,but, far from suspectingthat Dupin is the culprit
who has ravished it from him, knows nothingof it. Moreover,what he
is left with is far from insignificantfor what follows. We shall return
to what brought Dupin to inscribe a message on his counterfeitletter.
Whatever the case, the Minister,when he tries to make use of it, will
be able to read these words, writtenso that he may recognize Dupin's
462
BarbaraJohnson
hand: ". . . Un dessein si funestel S'il n'est digne d'Atrde est digne de
Thyeste,"6 whose source, Dupin tells us, is Crdbillon'sAtree.
Need we emphasize the similarityof these two sequences? Yes, for
the resemblance we have in mind is not a simple collection of traits
chosen only in order to delete their difference.And it would not be
enough to retain those common traits at the expense of the others for
the slightesttruthto result.It is ratherthe intersubjectivity in which the
two actions are motivated that we wish to bring into relief, as well
as the three terms throughwhich it structuresthem.
The special status of these terms results from their corresponding
simultaneouslyto the three logical momentsthroughwhich the decision
is precipitatedand the threeplaces it assigns to the subjects among whom
it constitutesa choice.
That decision is reached in a glance's time. For the maneuverswhich
follow, however stealthilythey prolong it, add nothing to that glance,
nor does the deferringof the deed in the second scene break the unity
of that moment.
This glance presupposes two others, which it embraces in its vision
of the breach left in their fallacious complementarity, anticipatingin it
the occasion for larceny affordedby that exposure. Thus three moments,
structuringthree glances, borne by three subjects, incarnated each time
by differentcharacters.
The firstis a glance that sees nothing: the King and the police.
The second, a glance which sees that the firstsees nothingand deludes
itself as to the secrecy of what it hides: the Queen, then the Minister.
The thirdsees that the firsttwo glances leave what should be hidden
exposed to whomeverwould seize it: the Ministerand finallyDupin.
In order to grasp in its unity the intersubjectivecomplex thus de-
scribed, we would willinglyseek a model in the technique legendarily
attributedto the ostrichattemptingto shield itselffromdanger; for that
technique might ultimatelybe qualified as political, divided as it here
is among three partners: the second believingitself invisible because the
firsthas its head stuck in the ground,and all the while lettingthe third
calmly pluck its rear; we need only enrich its proverbialdenomination
by a letter,producing la politique de l'autruiche,7 for the ostrich itself
to take on forevera new meaning.
Given the intersubjectivemodulus of the repetitiveaction, it remains
to recognize in it a repetitionautomatismin the sense that interestsus
in Freud's text. (SPL, pp. 41-44.)
6 "So infamous a scheme/ If not worthy of Atreus, is worthy of

Thyestes."
7 La politique de l'autruichecombines the policy of the ostrich (au-
truche),others(autrui) and Austri-,(Autriche).
463
Yale FrenchStudies
Thus,it is neitherthe characterof the individualsubjects,nor

the contentsof the letter,but thepositionof the letterwithinthe
group,whichdecideswhateach personwill do next.It is the fact
thatthe letterdoes not function as a unitof meaning(a signified)
but as thatwhichproducescertaineffects (a signifier),
whichleads
Lacan to read the storyas an illustration of "the truthwhichmay
be drawnfromthatmomentin Freud'sthought understudy-name-
ly,thatit is thesymbolicorderwhichis constitutive forthesubject
-by demonstrating (...) the decisiveorientationwhichthe subject
receivesfromthe itinerary of a signifier"(SPL, p. 40). The letter
acts like a signifier
preciselyto the extentthatits functionin the
storydoes not requirethat its meaningbe revealed: "the letter
was able to produceits effectswithinthe story: on the actorsin
the tale, includingthe narrator, as well as outsidethe story: on
us,thereaders, and also on itsauthor,withoutanyone'severbother-
ingto worryaboutwhatit meant"(not translatedin SPL; Ecrits,
and emphasismine)."The PurloinedLetter"thus
p. 57, translation
becomes for Lacan a kind of allegoryof the signifier.
Derrida'scritiqueofLacan'sreadingdoesnotdisputethevalidity
on its own terms,but questions
of the allegoricalinterpretation
and its modusoperandi.Derrida
ratherits implicitpresuppositions
aimshis objectionsat twokindsoftargets:1) whatLacan putsinto
the letterand 2) what Lacan leaves out of the text.
1) What Lacan puts into the letter.While asserting that the
letter'smeaningis lacking,Lacan,accordingto Derrida,makesthis
lackintothemeaningoftheletter.ButDerridadoes notstopthere:
he goes on to assertthatwhat Lacan meansby thatlack is the
truthof lack-as-castration-as-truth:"The truthof the purloined
letteris the truthitself(...) What is veiled/unveiledin this case
is a hole,a non-being thetruthofbeing[l'etre],as non-
[non-etant];
being.Truthis 'woman'as veiled/unveiled castration"(PT, pp. 60-
61). Lacan himself,however,neveruses the word "castration"in
thetextoftheoriginalSeminar.That it is suggestedis indisputable,
butDerrida,byfilling in whatLacanleftblank,is repeating precisely
the gestureof blank-fillingforwhichhe is criticizing
Lacan.
464
BarbaraJohnson
2) What Lacan leaves out of the text. This objection is itself

double: on the one hand,DerridacriticizesLacan forneglecting
to consider"The PurloinedLetter"in connectionwiththe other
two storiesin what Derridacalls Poe's "Dupin Trilogy."And on
the otherhand,accordingto Derrida,at the verymomentLacan
is readingthestoryas an allegoryof the signifier, he is beingblind
to thedisseminating powerofthesignifier inthetextoftheallegory,
in whatDerridacalls the "scene of writing." To cut out partof a
text'sframeof reference as thoughit did not exist,and to reduce
a complextextualfunctioning to a singlemeaning, are seriousblots
indeedin theannalsofliterary criticism.
Therefore it is all themore
noticeablethatDerrida'sownreadingofLacan'stextrepeatsprecise-
ly the crimesof whichhe accuses it: on the one hand,Derrida
makesno mentionof Lacan's longdevelopment on therelationbe-
tweensymbolicdetermination and randomseries.And on theother
hand,DerridadismissesLacan's "style"as a mereornament, veiling,
fora time,an unequivocalmessage: "Lacan's 'style',moreover, was
such thatfora long timeit wouldhinderand delayall access to
a unique contentor a single unequivocalmeaningdeterminable
beyondthewriting itself"(PT, p. 40). The factthatDerridarepeats
the verygestureshe is criticizing does not in itselfinvalidatehis
criticismof theireffects, but it does problematize his statement
condemning theirexistence.
Whatkindof logicis it thatthusseemsto turnone-upmanship
into inevitableone-downmanship?
It is preciselythe logic of the purloinedletter.
465
Yale FrenchStudies
3. Odd Couples
Je tiens la reineI
0 suir chatiment...
-Mallarme, "L'apres-midid'un faune"
L'ascendant que le ministretire de la situa-

tion ne tientdonc pas A la lettre,mais, qu'il
le sache ou non, au personnage qu'elle lui
constitue.
Lacan, "Seminairesur 'la Lettrevolee'e"
We have just seen how Derrida,in his effortto right(write)

Lacan's wrongs, can,on a certainlevel,onlyrepeatthem.And how
therectificationofa previousinjusticesomehowirresistibly dictates
thefillingin of a blank,whichthenbecomesthe new injustice.In
fact,the act of clinchingone's triumphby fillingin a blank is
alreadyprescribedin all its detailswithinPoe's story,in Dupin's
unwillingness blank"(Poe, p. 219) in thefac-
to "leave the interior
similehe has leftforthe Minister, in place of the purloinedletter
he, Dupin,has just repossessedby meansof a preciserepetition
of the act of robberyhe is undoing.Whatis writtenin the blank
is a quotation-as-signature, which curiouslyresemblesDerrida's
initialedinterventionsin the passageshe quotesfromLacan, a re-
semblanceon whichDerridais undoubtedly playing.And the text
ofthequotationtranscribed byDupinsaysprecisely thestructureof
whichhas led to its being
rectification-as-repetition-of-the-crime
transcribed in the firstplace:
-Un dessein si funeste,

S'il n'est digne d'Atree, est digne de Thyeste.
Atreus,whosewifewas longago seducedby Thyestes,is aboutto

makeThyesteseat (literally)the fruitof thatillicitunion,his son
Plisthenes.The avenger'splot may not be worthyof him,says
Atreus,but his brotherThyestesdeservesit. What the addressee
oftheviolenceis goingto getis simplyhis ownmessagebackwards.
angerwhich,as bothLacan and Derridashow,
It is thisvengeful
466
BarbaraJohnson
places Dupin as one of the "ostriches"in the "triad."Not content

simplyto returntheletterto its"rightful" Dupinjumps
destination,
intothe frayas the wrongedvictimhimself, by recallingan "evil
turn"the Ministeronce did him in Vienna,and forwhichhe is
now,personally, takinghis revenge.
Correctionmust thus posit a previouspretextual, pre-textual
crimewhichwill justifyits excesses.Any degreeof violenceis
permissible in theact ofgettingeven("To be evenwithhim,"says
Dupin,"I complainedof myweakeyes..." [Poe,p. 216]).And Du-
pin's backwardsrevisionof the storyrepeatsitselfin his readers
as well. The existenceof the same kind of prioraggressionon
Lacan's partis positedby Derridain a long footnotein his book
Positions,in whichhe outlineswhat will later develop into Le
Facteurde la Write:"In the textsI havepublishedup to now,the
absenceof reference to Lacan is indeedalmosttotal.That is justi-
fiednot onlyby theacts of aggressionin theformof,or withthe
intention of,reappropriation which,eversinceDe la grammatologie
appearedin Critique(1965)(and evenearlier,I am told)Lacan has
multiplied...."8 The priority of aggressionis doubledby the ag-
gressiveness of priority:"At the time of my firstpublications,
Lacan's tcritshad not yetbeen collectedand published..." 9 And
Lacan, in turn,mentionsin his Presentation to the Pointsedition
of his tcrits:"whatI properly call the instanceof theletterbefore
anygrammatology." 10The rivalryoversomething neithermanwill
creditthe otherwithpossessing,the retrospective revisionof the
originsofboththeirresemblances thusspirals
and theirdifferences,
backwardand forwardin an indeterminable patternof cancellation
and duplica~tion. If it thusbecomesimpossibleto determine "who
startedit" (or evenwhether "it" was startedbyeitherone ofthem),
it is also impossibleto knowwho is ahead,or evenwhose"turn"
it is. This is whatmakesthebusinessof "getting even"so odd.
8 J. Derrida, Positions, Minuit, 1972, pp. 112-113 (translationand em-

phasis mine).
9 Ibid., p. 113.
10 J. Lacan, Ecrits, Seuil ("Points"), 1966, pp. 11 (translationand em-
phasis mine).
467
Yale French Studies
This type of oscillation between two terms, considered as

totalitiesin a binaryopposition,is studied by Lacan in connection
with Poe's story of the eight-year-oldprodigywho succeeded in
winning,far beyond his due, at the game of-even and odd. The
game consists of guessing whetherthe number of marbles an op-
ponentis holdingis even or odd. The schoolboyexplains his success
by his identificationwiththe physicalcharacteristicsof his opponent,
fromwhich he deduces the opponent's degree of intelligenceand
its correspondingline of reasoning.What Lacan shows, in the part
of his seminar which Derrida neglects,is that the mere identifica-
tion with the opponent as an image of totalityis not sufficientto
insure success-and in no way explains Dupin's actual strategy-
since, fromthe momentthe opponentbecomes aware of it, he can
then play on his own appearance,and dissociate it fromthe reason-
ing that is presumed to go with it. (This is, indeed, what occurs
in the encounterbetween Dupin and the Minister: the Minister's
feignednonchalanceis a true vigilancebut a blinded vision,whereas
Dupin's feignedblindness("weak eyes") is a vigilantact of lucidity,
later to succumb to its own formof blindness.)From then on, says
Lacan, the reasoning "can only repeat itself in an indefiniteoscil-
lation" (?crits, p. 58: translationmine). And Lacan reportsthat,in
his own classroomtests of the schoolboy'stechnique,it was almost
inevitable that each player begin to feel he was losing his
marbles...
.1
But if the complexities of these texts could be reduced to a
mere combat between ostriches,a mere game of heads and tails
played out in order to determinea "winner," they would have
very little theoretical interest.It is, on the contrary,the way in
which each mastermindavoids simply becoming the butt of his
own joke that displaces the oppositionin unpredictableways, and
transformsthe textual encounter into a source of insight.For if
the verypossibilityof meetingthe opponenton a common ground,
withoutwhich no contact is possible, implies a certain symmetry,
11 Cf. Lacan's description of the "effect of disorientation, or even of

great anxiety," provoked by these exercises (Ecrits, p. 60).
468
Barbara Johnson
a sameness,a repetitionof the errorwhichthe encounteris designed

to correct,any true avoidance of that error entails a non-meeting
or incompatibilitybetween the two forces. If to hit the target is
in a way to become the target,then to miss the targetis perhaps
to hit it elsewhere.It is not the way in which Lacan and Derrida
meet each other,but ratherthe way in which they miss each other,
which opens up a space for interpretation.
Clearly, what is at stake here has somethingto do with the
status of the number 2. If the face-offbetween two opponents or
polar opposites always simultaneouslybackfiresand misfires,it can
only be because 2 is an extremelyodd number.On the one hand,
as a specular illusion of symmetryor metaphor,it can either be
narcissisticallyreassuring(the image of the other as a reinforce-
ment of my identity)or absolutely devastating(the other whose
existence can totally cancel me out). This is what Lacan calls the
imaginaryduality. It is characterizedby its absoluteness, its in-
dependence fromany accident or contingencywhich mightsubvert
the unityof the terms in question, whetherin their oppositionor
in their fusion. To this, Lacan opposes the symbolic,which is the
entrance of differenceor otherness or temporalityinto the idea
of identity-it is not somethingwhich befallsthe imaginaryduality,
but somethingwhich has always already inhabited it, something
which subvertsnot the symmetryof the imaginarycouple, but the
possibilityof the independentunityof any one termwhatsoever.It
is the impossibilitynot of the number2 but of the number1. Which,
paradoxicallyenough, turns out to lead to the number 3.
If 3 is what makes 2 into the impossibilityof 1, is there any
inherentincrease in lucidityin passing froma couple to a triangle?
Is a trianglein any way more "true" than a couple?
It is Derrida's contentionthat, for psychoanalysis,the answer
to that question is "yes." The trianglebecomes the magical,Oedipal
figurethat explains the functioningof human desire. The child's
originalimaginarydual unitywith the motheris subvertedby the
law of the father as that which prohibitsincest under threat of
castration,and the child has "simply" to "assume castration" as
469
Yale FrenchStudies
the necessityof substitution in the objectof his desire(the object

of desirebecomingthe locus of substitution and the focusof re-
petition),afterwhich the child's desire becomes "normalized."
Derrida'scriticism of the "triangles"or "triads"in Lacan's reading
of Poe is based on theassumption thatLacan's use of triangularity
stemsfromthispsychoanalytical myth.
Derrida'scriticismtakes two routes,both of themnumerical:
1) The structure of "The PurloinedLetter"cannotbe reduced
to a triangleunlessthe narratoris eliminated.The elimination of
the narratoris a blatantand highlyrevealingresultof the way
"psychoanalysis"does violenceto literature
in orderto finditsown
schemes.Whatpsychoanalysis sees as a triangleis therefore
really
a quadrangle,and thatfourthside is thepointfromwhichliterature
problematizes the verypossibilityof a triangle.Therefore:3 = 4.
2) Dualityas suchcannotbe dismissedor simplyabsorbedinto
a triangular structure."The PurloinedLetter"is traversedby an
uncannycapacityfordoublingand subdividing. The narratorand
Dupin are doublesof each other,and Dupin himselfis firstin-
troducedas a "Bi-PartSoul" (Poe, p. 107),a sortof DupinDuplex,
"the creativeand the resolvent." The Minister,D-, has a brother
forwhomit is possibleto mistakehim,and fromwhomhe is to be
distinguished because of his doubleness(poetand mathematician).
Thus theMinisterand Dupinbecomedoublesofeach otherthrough
the factof theirboth beingalreadydouble,in additionto their
otherpointsof resemblance, includingtheirnames."The Seminar,"
writesDerrida,"mercilessly forecloses
thisproblematic ofthedouble
and of Unheimlichkeit-no doubtconsidering thatit is confined to
theimaginary, to thedual relationship
whichmustbe keptrigorous-
ly separatefromthe symbolicand the triangular. (...) All the
'uncanny'relationsofduplicity, deployedin a dual struc-
limitlessly
ture,findthemselves omittedor marginalized[in theSeminar].(...)
What is thuskept undersurveillance and controlis the Uncanny
itself,and thefranticanxietywhichcan be provoked, withno hope
of reappropriation, enclosure,or truth,by the infinite play from
470
BarbaraJohnson
simulacrum to simulacrum,fromdoubleto double"(omittedin PT;

FV, p. 124,translation mine).
Thus the triangle's
anglesare alwaysalreadybisected,and 3 =
(a factorof) 2.
In thegameofodd versuseven,then,it wouldseemthatDerrida
is playingevens(4 or 2) againstLacan's odds (3). But somehowthe
numbers2 and 4 have becomeuncannilyodd, whilethe number3
has been evenedoffintoa reassuring symmetry. How did thishap-
pen,and whatare the consequencesforan interpretation of "The
PurloinedLetter"?
Beforeany answerto this questioncan be envisaged,several
remarks shouldbe madehereto problematize thetermsofDerrida's
critique:
1) If the narratorand Dupin are a strictlydual pair whose
relationshipis in no way mediatedby a thirdtermin anyOedipal
sense,how is one to explainthe factthattheiroriginalmeeting
was broughtaboutby theirpotentialrivalryoverthesame object:
"theaccidentof ourbothbeingin searchofthesameveryrareand
veryremarkable volume."Whetheror not theyever foundit, or
can shareit,is thisnota triangularrelationship?
2) AlthoughLacan's readingof "The PurloinedLetter"divides
the storyinto triadicstructures,his modelfor(inter-)subjectivity,
the so-called"schemaL," whichis developedin thatpart of the
Seminar'sintroduction whichDerridaglossesover,is indisputably
quadrangular. In orderto read Lacan's repeatingtriadsas a trian-
gular,Oedipalmodel of the subjectinsteadof as a merestructure
of repetition, Derrida must therefore lop offone cornerof the
"schemaL" in the same way as he accusesLacan of loppingoffa
cornerof Poe's text-and Derridadoes this preciselyby lopping
offthatcornerofLacan'stextin whichthequadrangular "schemaL"
is developed.
But can what is at stake here reallybe reducedto a mere
numbersgame?
Let us approachtheproblemfromanotherangle,by askingtwo
more questions:
471
Yale FrenchStudies
1) Whatis therelationbetweena dividedunityand a duality?

Are the two two's synonymous? Is a "Bi-PartSoul," forexample,
actuallycomposedof two wholes?Or is it possibleto conceiveof
a divisionwhichwould not lead to two separableparts,but only
of the idea of unity?This wouldclass what
to a problematization
Derridacalls "duality"not in Lacan's "imaginary,"
but in Lacan's
"symbolic."
2) If the doubles are foreverredividing or multiplying, does
the number"2" reallyapply?If 1 = 2, how can 2 = 1 + 1? If
whatis uncannyaboutthedoublesis thattheyneverstopdoubling
up, wouldthe number2 stillbe uncannyif it did stop at a truly
dual symmetry? Isn'tit theverylimitlessness of theprocessof the
dissemination ofunity, rather than the existenceofanyone duality,
whichDerridais talkingabouthere?
Clearly,in thesequestions,it is the verynotionof a number
whichbecomesproblematic, and the argumenton the basis of
numbers can no longerbe readliterally. If Derridaopposesdoubled
quadrangles it is notbecausehe wantsto turn
to Lacan's triangles,
Oedipusintoan octopus.
To what,then,does the critiqueof triangularity apply?
The problemwith psychoanalytical triangularity, in Derrida's
eyes,is notthatit containsthewrongnumberofterms,but thatit
presupposes thepossibilityof a successfuldialecticalmediationand
harmonious normalization or Aufhebung of desire.The threeterms
in the Oedipaltriadenterintoan oppositionwhoseresolutionre-
semblesthe synthetic momentof a Hegeliandialectic.The process
centerson the phallusas the locus of the questionof sexual dif-
ference:it is whenthe observation of the mother'slack of a penis
is joined withthe father'sthreatof castrationas the punishment
forincestthatthechildpassesfromthealternative (thesisvs. anti-
thesis; presencevs. absenceof penis)to the synthesis (the phallus
as a signof the factthatthe childcan onlyenterintothe circuit
of desireby assumingcastrationas the phallus'simultaneous pres-
ence and absence; that is, by assumingthe fact that both the
subjectand theobjectof desirewillalwaysbe substitutes forsome-
472
BarbaraJohnson
thingthat was never reallypresent).In Lacan's articleon "La

signification
du phallus,"which Derrida quotes, this process is
evokedin preciselyHegelianterms:
All these remarksstill do nothingbut veil the fact that it [the phallus]
cannot play its role exceptveiled, that is to say as itselfsign of the latency
with which anythingsignifiableis strickenas soon as it is raised (aufge-
hoben) to the functionof signifier.
The phallus is the signifierof this Aufhebungitselfwhich it inaugurates
(initiates)by its disappearance.(l6crits,p. 692; PT, p. 98.)
"It would appear,"commentsDerrida,"thatthe Hegelianmove-

mentof Authebungis here reversedsince the lattersublates[re-
leve] the sensorysignifier
in the ideal signified"
(PT, p. 98). But
then,accordingto Derrida,Lacan's privilegingof the spokenover
thewritten wordannullsthisreversal,reappropriatesall possibility
of uncontainable otherness,
and bringsthe wholethingback within
theboundsof thetypeof "logocentrism" whichhas been thefocus
ofDerrida'sentiredeconstructiveenterpriseoverthepasttenyears.
The questionof whetheror not Lacan's "privileging" of the
voice is strictly in Derrida'ssenseis an extremely
logocentric com-
plexone withwhichwe cannothopeto deal adequatelyhere.12 But
whatdoes all thishave to do with"The PurloinedLetter"?
In an attemptto answerthisquestion,let us examinethe way
in whichDerridadeducesfromLacan'stextthefactthat,forLacan,
the"letter"is a symbolofthe(mother's) phallus.SinceLacan never
uses the word "phallus"in the Seminar,this is alreadyan inter-
12 Some idea of the possibilitiesfor misunderstanding inherentin this

question can be gatheredfromthe following:In orderto show thatpsycho-
analysis represses"writing"in a logocentricway, Derrida quotes Lacan's
statementagainst tape recorders: "But preciselybecause it comes to him
throughan alienated form,even a retransmission of his own recordeddis-
course, be it from the mouth of his own doctor, cannot have the same
effectsas psychoanalyticalinterlocution."This Derrida regards as a con-
demnationof the "simulacrum,"a "disqualificationof recordingor of re-
petitionin the name of the livingand presentword." But what does Lacan
actually say? Simplythat a tape recordingdoes not have the same effects
as psychoanalyticalinterlocution.Does the fact that psychoanalysisis a
techniquebased on verbal interlocutionautomaticallyreduce it to a logo-
centric error? Is it not equally possible to regard what Lacan calls "full
speech" as being full of preciselywhat Derrida calls writing?
473
Yale FrenchStudies
pretationon Derrida'spart,and quite an astuteone at that,with

whichLacan,as a laterreaderof his ownSeminar,implicitly agrees
by placingthe word"castrated"-which had not been used in the
originaltext-in his "Points"Presentation.The disagreement be-
tweenDerridaand Lacan thusarisesnot over the validityof the
equation"letter= phallus,"but overits meaning.
How,then,does DerridaderivethisequationfromLacan'stext?
The deductionfollowsfourbasic lines of reasoning,all of which
willbe dealtwithin greaterdetaillaterin the presentessay:
1) The letter"belongs"to the Queen as a substituteforthe
phallusshe does not have. It feminizes
(castrates)each of its suc-
cessiveholdersand is eventuallyreturnedto her as its rightful
owner.
2) Poe's description
ofthepositionoftheletterintheMinister's
apartment,expandeduponby the figurative dimensionsof Lacan's
text,suggestsan analogybetweenthe shapeof the fireplacefrom
thecenterofwhosemantelpiece theletteris foundhanging,
and that
pointon a woman'sanatomyfromwhichthephallusis missing.
3) The letter,saysLacan,cannotbe divided: "But ifit is first
of all on the materialityof the signifierthatwe have insisted,that
materiality is odd [singuliere]in manyways,the firstof whichis
notto admitpartition" (SPL, p. 53). Thisindivisibility,
saysDerrida,
is odd indeed,butbecomescomprehensible ifit is seenas an ideal-
izationof the phallus,whoseintegrity is necessaryforthe edifica-
tionof the entirepsychoanalytical system.Withthe phallussafely
idealizedand locatedin thevoice,the so-called"signifier" acquires
the "unique,living,non-mutilable integrity"of the self-present
spokenword,unequivocally pinneddownto and by the signified.
"Had the phallusbeen per(mal)chance divisibleor reducedto the
statusofa partialobject,thewholeedification wouldhavecrumbled
down, and this is what has to be avoided at all cost" (PT,
pp. 96-97).
4) And finally,
if Poe's story"illustrates"
the "truth,"the last
wordsof the Seminarproperseem to reaffirm that truthin no
474
BarbaraJohnson
uncertainterms: "Thus it is thatwhatthe 'purloinedletter',nay,

the 'letterin sufferance'meansis thata letteralwaysarrivesat its
destination" (SPL, p. 72, emphasismine).Now, sinceit is unlikely
thatLacan is talkingabout the efficiency of the postalservice,he
must,accordingto Derrida,be affirming thepossibility of unequiv-
ocal meaning, the eventualreappropriation of themessage,its total
equivalencewith itself.And since the "truth"Poe's storyillus-
tratesis, in Derrida'seyes,the truthof veiled/unveiled castration
and of the transcendental identityof the phallusas the lack that
makes the systemwork,this finalsentencein Lacan's Seminar
seemsto affirm boththeabsolutetruthofpsychoanalytical theories
and theabsolutedecipherability of theliterary text.Poe's message
will have been totally,unequivocally understood and explainedby
thepsychoanalytical myth."The hermeneutic discovery of meaning
(truth),the deciphering (that of Dupin and thatof the Seminar),
arrivesitselfat its destination" (PT, p. 66).
Thus, the law of the phallusseemsto implya reappropriating
returnto the place of trueownership, an indivisibleidentityfunc-
tioningbeyondthe possibilityof disintegration or unrecoverable
loss, and a totallyself-present, unequivocalmeaningor truth.
The problemwiththistypeof system, countersDerrida,is that
it cannotaccountforthe possibility of sheeraccident,irreversible
loss, unreappropriable residues,and infinite divisibility,
whichare
in factnecessary and inevitable in thesystem'sveryelaboration. In
orderforthe circuitof the letterto end up confirming the law
of the phallus,it mustbeginby transgressing it: the letteris a
sign of high treason.Phallogocentrism mercilessly repressesthe
uncontrollable multiplicity of ambiguities, the disseminating play
of writing, which irreduciblytransgressesany unequivocal meaning.
"Not thattheletterneverarrivesat its destination,
but partof its
structure
is thatit is alwayscapableofnotarrivingthere.(...) Here
dissemination and of castration
threatensthe law of the signifier
as a contractof truth.Dissemination mutilatesthe unityof the
signifier,
thatis, of the phallus"(PT, p. 66).
475
Yale FrenchStudies
In contrastto Lacan'sSeminar,then,Derrida'stextwouldseem
to be settingitselfup as a Disseminar.
Fromthe foregoing remarks, it can easilybe seen thatthe dis-
seminalcriticism of Lacan's apparentreduction of the literary
text
to an unequivocalmessagedependsforits forceupon the presup-
positionof unambiguousness in Lacan'stext.And indeed,thestate-
mentthat a letteralwaysreachesits destinationseems straight-
forwardenough.But when that statementis reinserted into its
context,thingsbecomepalpablyless certain:
Is thatall, and shall we believe we have decipheredDupin's real strategy

above and beyond the imaginarytricks with which he was obliged to
deceive us? No doubt, yes, for if "any point requiringreflection,"as
Dupin states at the start,is "examined to best purpose in the dark," we
may now easily read its solution in broad daylight.It was already implicit
and easy to derivefromthe title of our tale, accordingto the veryformula
we have long submittedto your discretion: in which the sender,we tell
you, receives fromthe receiverhis own message in reverseform.Thus it
is that what the "purloined letter,"nay, the "letter in sufferance"means
is that a letteralways arrivesat its destination.(SPL, p. 72.)
The meaningof thislast sentenceis problematized notso much

by its own ambiguity as by a seriesof reversalsin the preceding
sentences.If the best examination takes place in darkness,what
does "readingin broaddaylight"imply?Could it not be takenas
an affirmation not of actual luciditybut of delusionsof lucidity?
Couldit notthenmovethe"yes,no doubt"as an answernotto the
question"have we deciphered?"but to the question"shall we
believewe have deciphered?"And if this is possible,does it not
emptythefinalaffirmation of all unequivocality,
leavingit to stand
withtheforceof an assertion, withoutanydefinite content?And if
the senderreceivesfromthe receiverhis own messagebackwards,
whois thesenderhere,whothereceiver, and whatis themessage?
It is in factnotevenclearwhattheexpression "thepurloinedletter"
refersto: Poe's text?the letterit talksabout? or simplythe ex-
pressiion"thepurloined letter"?
We will take anotherlook at this passage later,but for the
momentits ambiguitiesseem sufficient to problematize, if not
476
BarbaraJohnson
subvert, thepresupposition of univocality whichis theveryfounda-

tionon whichDerridahas edifiedhis interpretation.
But surelysuchan oversimplification on Derrida'spartdoes not
resultfrommereblindness, oversight, or error.As P. de Man says
of Derrida'ssimilartreatment of Rousseau,"the patternis too
interesting notto be deliberate." 13 Beingthesharp-eyed readerthat
he is, Derrida'sconsistentforcing ofLacan'sstatements intosystems
and patternsfromwhichtheyare actuallytryingto escape must
correspondto some strategicnecessitydifferent fromthe atten-
tivenessto the letterof the textwhichcharacterizes Derrida'sway
of readingPoe. And in fact,the moreone workswithDerrida's
analysis,themoreconvincedonebecomesthatalthoughthecritique
ofwhatDerridacalls psychoanalysis is entirely
justified,it does not
quite applyto whatLacan's textis actuallysaying.What Derrida
is in factarguingagainstis therefore not Lacan's textbut Lacan's
power-or rather,"Lacan" as the apparentcause of certaineffects
ofpowerin Frenchdiscoursetoday.Whatever Lacan'stextmaysay,
it functions, accordingto Derrida,as ifit said whathe saysit says.
The statement thata letteralwaysreachesits destination maybe
totallyundecipherable, but its assertiveforceis takenall the more
seriously as a signthatLacan himself has everythingall figured out.
Such an assertion, of
in fact,giveshiman appearance mastery like
thatoftheMinisterin theeyesoftheletterless Queen."The ascen-
dancy which the Ministerderivesfromthe situation,"explains
Lacan, "is attachednot to theletterbut to the characterit makes
himinto."
Thus Derrida'sseemingly"blind" reading,whose vagarieswe
shall be followinghere,is not a mistake,but the positioning of
whatcan be calledthe "averagereading"of Lacan's text,whichis
the trueobject of Derrida'sdeconstruction. Since Lacan's text is
read as if it said whatDerridasays it says,its actualtextualfunc-
tioningis irrelevant to the agonisticarenain whichDerrida'sanal-
13 Paul de Man, Blindness and Insight,Oxford UniversityPress, 1971,

P. 140.
477
Yale FrenchStudies
ysis takesplace,and whichis, in fact,suggestedby the veryfirst

wordof the epigraph:ils, "they":
Theythankhimforthegrandtruths he has justproclaimed,

- forthey
have discovered(o verifierof whatcannotbe verified!) that everything
he said was absolutely
true; eventhough, thesehonestsoulsadmit,
at first,
theymighthave suspectedthat it could have been a simplefiction...
(PT, p. 31; translation
mine.)
The factthatthisquotationfromBaudelairerefersto Poe and not

Lacan does notcompletely erasetheimpression thattheunidentified
"him"in its firstsentenceis the "Purveyorof Truth"of the title.
The evils of Lacan's analysisof Poe are thus locatedless in the
letterof the text than in the gulliblereaders,the "bravesgens"
who are takenin by it. Lacan's ills are reallyius.
If Derrida'sreadingof Lacan's readingof Poe is thus actually
the deconstructionof a reading whose status is difficultto deter-
mine,does this mean thatLacan's textis completely innocentof
the misdemeanors of whichit is accused? If Lacan can be shown
to be opposedto thesamekindof logocentric errorthatDerridais
opposedto, does thatmean thattheyare both reallysayingthe
same thing?These are questionswhichmustbe left,at least for
the moment, hanging.
ButthestructureofDerrida'stransference ofguiltfroma certain
readingof Lacan onto Lacan's textis not indifferentin itself,in
the contextof what,afterall, startedout as a relativelysimple
crime story. For what it amounts to is nothingless than ... a
frame.
4. The Frameof Reference

nue en le miroir,
Elle,ddfunte encor
Que, dans l'oubli fermepar le cadre, se fixe
De scintillationssitot le septuor.
-Mallarm., "Sonneten X"
If Derrida is thus framingLacan foran interpretative

malpractice
of which he himselfis, at least in part, the author,what can this
478
BarbaraJohnson
frameteachus aboutthenatureoftheact ofreading, in thecontext

of the questionof literatureand psychoanalysis?
Interestingly
enough,one of the majorcrimesforwhichLacan
is beingframed byDerridais preciselythepsychoanalytical
reading's
elimination of the literarytext'sframe.That framehere consists
not onlyof the two storieswhichprecede"The PurloinedLetter,"
but of the stratumof narration throughwhichthe storiesare told,
and, "beyond" it, of the text's entirefunctioningas &riture:
Withoutbreathinga word about it, Lacan excludes the textual fiction

withinwhichhe isolates the so-called "generalnarration."Such an operation
is facilitated,too obviouslyfacilitated,by the fact that the narrationcovers
the entiresurface of the fictionentitled"The Purloined Letter." But that
is the fiction.There is an invisiblebut structurally irreducibleframearound
the narration.Where does it begin? With the firstletterof the title? With
the epigraphfromSeneca? With the words,"At Paris, just afterdark..."?
It is more complicated than that and will require reconsideration.Such
complicationsufficesto point out everythingthat is misunderstoodabout
the structureof the text once the frameis ignored. Within this invisible
or neutralized frame, Lacan takes the borderless narration and makes
anothersubdivision,once again leaving aside the frame.He cuts out two
dialogues from within the frame of the narrationitself,which form the
narratedhistory,i.e. the contentof a representation, the internalmeaning
of a story,the all-enframed whichdemandsour completeattention,mobilizes
all the psychoanalyticalschemes-Oedipal, as it happens-and draws all
the effortof deciphermenttowards its center. What is missinghere is an
elaboration of the problem of the frame,the signatureand the parergon.
This lack allows us to reconstructthe scene of the signifieras a signified
(an ever inevitableprocess in the logic of the sign), writingas the written,
the text as discourse or more preciselyas an "intersubjective"dialogue
(thereis nothingfortuitousin the fact that the Seminardiscusses only the
two dialogues in "The Purloined Letter"). (PT, pp. 52-53, translation
modified.)
It is well knownthat "The PurloinedLetter"belongsto what Baudelaire
called a "kind of trilogy,"along with "The Murders in the Rue Morgue"
and "The Mysteryof Marie Roget." About this Dupin trilogy,the Seminar
does not breathea word; not onlydoes Lacan liftout the narratedtriangles
(the "real drama") in order to centerthe narrationaround them and make
them carry the weight of the interpretation (the letter's destination),but
he also liftsone thirdof the Dupin cycle out of an ensemblediscarded as
if it were a natural,invisible frame.(Not translatedin PT; FV p. 123;
translationmine.)
479
Yale FrenchStudies
In framingwith such violence,in cuttinga fourthside out of the nar-

rated figureitself in order to see only triangles,a certain complication,
perhaps a complicationof the Oedipal structure,is eluded, a complication
which makes itself felt in the scene of writing.(PT, p. 54; translation
entirelymodified.)
It would seem,then,thatLacan is here guiltyof severalsins

of omission: the omissionof the narrator,of the non-dialogue
partsof thestory,of the otherstoriesin thetrilogy. But does this
criticism amountto a mereplea fortheinclusionof whathas been
excluded?No: the problemis not simplyquantitative. What has
been excludedis not homogeneousto what has been included.
Lacan, says Derrida,missesthe specifically literarydimensionof
Poe's textby treating it as a "real drama,"a storylike the stories
a psychoanalyst hearseveryday fromhis patients.Whathas been
leftout is precisely
literatureitself.
Does thismeanthatthe "frame"is whatmakesa textliterary?
Interestinglyenough,in a recentissue of New LiteraryHistory
devotedto thequestion"Whatis Literature?" and totallyunrelated
to the debateconcerning the purloinedletter,thisis preciselythe
conclusionto whichone of the contributors comes: "Literature
is language (...), but it is language around which we have drawn
a frame,a framethatindicatesa decisionto regardwitha particular
self-consciousness
the resourceslanguagehas alwayspossessed."14
Sucha viewofliterature,however, impliesthata textis literary
because it remainsinsidecertaindefiniteborders: it is a many-
facettedobject,perhaps,but still,it is an object.That thisis not
quite whatDerridahas in mindbecomesclear fromthe following
remarks:
By overlookingthe narrator'sposition, the narrator'sinvolvementin

the contentof what he seems to be recounting,one omits fromthe scene
of writinganythinggoing beyond the two triangularscenes.
And firstof all one omits that what is in question-with no possible
access route or border-is a scene of writingwhose boundaries crumble
offinto an abyss.From the simulacrumof an overture,of a "firstword,"the
14 StanleyE. Fish, "How Ordinaryis OrdinaryLanguage?,"New Literary

History, Vol. V, No. 1, p. 52 (emphasismine).
480
BarbaraJohnson
narrator,in narratinghimself,advances a few propositionswhich carry

the unityof the "tale" into an endless drifting-off
course: a textualdrifting
not at all taken into account in the Seminar.(PT, pp. 100-101; translation
modified.)
These reminders,of which countless other examples could be given,
alert us to the effectsof the frame,and of the paradoxes in the parergonal
logic. Our purpose is not to prove that "The Purloined Letter" functions
withina frame(omittedby the Seminar,which can thus be assured of its
triangularinteriorby an active, surreptitiouslimitationstartingfrom a
metalinguisticoverview),but to prove that the structureof the framing
effectsis such that no totalizationof the border is even possible. Frames
are always framed: thus,by part of theircontent.Pieces withouta whole,
"divisions" withouta totality-this is what thwartsthe dream of a letter
without division, allergic to division. (PT, p. 99; translation slightly
modified.)
Here theargument seemsto reversethepreviousobjection:Lacan

has eliminatednot the framebut the Vnframability of the literary
text.But whatDerridacalls "parergonal logic" is paradoxicalpre-
ciselybecause both of theseincompatible (but not totallycontra-
dictory)argumentsare equally valid. The total inclusionof the
"frame"is both mandatory and impossible.The "frame"thusbe-
comesnot the borderline betweenthe insideand the outside,but
preciselywhatsubvertstheapplicability of theinside/outside polar-
ityto the act of interpretation.
The "frame"is, in fact,one ofa seriesofparadoxical"borderline
cases"-along withthe tympanum and the hymen-through which
Derridahas recently been studying the limitsof spatiallogicas it
relatesto intelligibility.
Lacan, too, has been seekingto displace
theEuclideanmodelofunderstanding (comprehension, forexample,
meansspatialinclusion)by inventing a "new geometry" by means
of the logicof knots.The relationbetweenthesetwo attemptsto
breakoutofspatiallogichas yetto be articulated, butsomemeasure
of the difficulties
involvedmaybe derivedfromthe factthat"to
break out of" is still a spatial metaphor.The urgencyof these
undertakings cannot,however,be overestimated, sincethelogicof
metaphysics, of politics,of belief,and of knowledgeitselfis based
on theimposition ofdefinable objectivefrontiers and outlineswhose
481
Yale FrenchStudies
possibility and/orjustifiability
are preciselywhatis herebeingput
in question.If "comprehension" is the framing of something whose
limitsare undeterminable, how can we know what we are com-
prehending? The playon the spatialand thecriminalsensesof the
word"frame"withwhichwe beganthissectionmaythusnotbe as
gratuitous as it seemed.And indeed,the questionof the fallacies
inherentin a Euclideanmodelof intelligibility, far frombeinga
tangential theoreticalconsideration here,is in factcentralto the
very plot of "The PurloinedLetter" itself.For it is precisely
thenotionof space as finiteand homogeneous whichunderliesthe
Prefect'smethodof investigation:"I presumeyou know," he
explains,"that,to a properly trainedpolice-agent, sucha thingas a
'secret'draweris impossible.Any man is a dolt who permitsa
'secret'drawerto escape him in a searchof this kind.The thing
is so plain. There is a certainamountof bulk-of space-to be
accountedforin everycabinet.Then we have accuraterules.The
fiftiethpartof a line could not escape us" (Poe, p. 204). The as-
sumptionthat what is not seen mustbe hidden-an assumption
Lacan callsthe"realist'simbecillity"-is basedon a falselyobjective
notionof the act of seeing.The polarity"hidden/exposed" cannot
alone accountforthe factthatthe police did not findthe letter
-which was entirelyexposed,insideout-let alone the factthat
Dupindid. A "subjective"elementmustbe added,whichsubverts
the geometrical model of understanding throughthe interference
ofthepolarity "blindness/sight"
withthepolarity "hidden/exposed."
The same problematic is raisedby the storyof "The Emperor's
New Clothes,"whichDerridacitesas an exampleofpsychoanalysis'
failureto go beyondthe polarity"hidden/exposed" (in Freud's
account).We willreturn to theletter's"place"lateron in thisessay,
butit is alreadyclearthatthe"range"ofanyinvestigation is located
not in geometricalspace, but in its implicitnotion of what
"seeing"is.
WhatenablesDerridato problematize the literarytext'sframe
is, as we have seen,whathe calls "the sceneof writing." By this
he meanstwothings:
482
BarbaraJohnson
1) The textualsignifier's resistanceto beingtotallytransformed

intoa signified.In spiteof Lacan's attentiveness to thepathof the
letterinPoe's storyas an illustration
ofthefunctioning ofa signifier,
says Derrida,the psychoanalytical readingis still blind to the
functioning of the signifierin the narrationitself.In reading"The
PurloinedLetter"as an allegoryof the signifier, Lacan, according
to Derrida,has made the "signifier" intothe story'struth: "The
displacement of the signifieris analyzedas a signified, as the re-
countedobject in a shortstory"(PT, p. 48). Whereas,counters
Derrida,it is preciselythetextualsignifier whichresistsbeingthus
totalizedinto meaning,leavingan irreducible residue: "The rest,
the remnant, wouldbe 'The PurloinedLetter',the textthatbears
thistitle,and whoseplace,liketheoncemoreinvisiblelargeletters
on themap,is notwhereone was expecting to findit,intheenclosed
contentof the 'real drama'or in the hiddenand sealed interior
of Poe's story,but in and as the open letter,the veryopen letter
whichfictionis" (PT, p. 64).
2) The actual writings-thebooks, libraries,quotations,and
previoustaleswhichsurround"The PurloinedLetter"witha frame
of (literary)
references.
The storybeginsin "a littlebacklibrary, or
book-closet"(Poe, p. 199) wherethe narratoris mullingover a
previousconversation on the subjectof the twopreviousinstances
of Dupin's detectiveworkas told in Poe's two previoustales,the
firstof whichrecountedthe originalmeetingbetweenDupin and
the narrator-ina library,of course,whereboth were in search
of the same rarebook. The story'sbeginning is thusan infinitely
regressing referenceto previouswritings.And therefore, says Der-
rida, "nothingbegins.Simplya drifting or a disorientation from
whichone nevermovesaway"(PT, p. 101).Dupinhimself is in fact
a walkinglibrary:booksare his "sole luxuries," and thenarrator is
"astonished"at "thevastextentof his reading"(Poe, p. 106).Even
Dupin's last, mostseemingly personalwords-the venomouslines
he leaves in his substitute
letterto the Minister-area quotation.
A quotationwhosetranscription and properauthorship are thelast
thingsthe storytells us. "But," concludesDerrida,"beyondthe
483
Yale FrenchStudies
quotationmarksthat surroundthe entirestory,Dupin is obliged

to quotethislast wordin quotationmarks,to recounthis signature:
thatis whatI wroteto himand howI signedit.Whatis a signature
withinquotationmarks?Then,withinthesequotationmarks,the
seal itselfis a quotationwithinquotationmarks.This remnantis
stillliterature" (PT, pp. 112-113).
It is bymeansofthesetwoextradimensions thatDerridaintends
to showthe crumbling, edgesof the story's
abyssal,non-totalizable
frame.Both of theseobjections,however,are in themselvesmore
problematic and double-edged thantheyappear.Let us beginwith
the second. "Literature,"in Derrida's demonstration, is indeed
clearlythe beginning, middle,and end-and even the interior-of
the purloinedletter.But how was this conclusionreached?To a
large extent,by listingthe books, libraries,and otherwritings
recountedin the story.That is, by following the theme-and not
the functioning-of "writing"within"the contentof a representa-
tion."But ifthefactthatDupinsignswitha quotation, forexample,
is forDerridaa sign that "this remnantis still literature," does
thisnot indicatethat"literature" has becomenot the signifier but
thesignified in thestory?If theplayof thesignifier is reallyto be
followed,doesn'tit play beyondthe rangeof the seme "writing"?
And if DerridacriticizesLacan formakingthe "signifier" intothe
story'ssignified, is Derridanot here transforming "writing"into
"the written"in much the same way? What Derridacalls "the
reconstruction of the scene of the signifier
as a signified" seems
indeedto be "an inevitableprocess"in the logic of readingthe
purloinedletter.
Derrida,of course,implicitly countersthisobjectionby protest-
ing-twice-thatthe textualdrifting forwhichLacan does not ac-
countshouldnot be considered"the real subjectof the tale,"but
ratherthe "remarkable ellipsis"of anysubject.But thequestionof
the seemingly inevitableslippingfromthe signifierto the signified
stillremains.And it remainsnotas an objectionto thelogicof the
frame,but as its fundamental question.For if the "paradoxesof
parergonal logic" are such thatthe frameis alwaysbeingframed
484
BarbaraJohnson
by partof its contents, it is preciselythisslippagebetweensignifier

and signified-which is actedoutbybothDerridaand Lacan against
theirintentions-which best illustratesthoseparadoxes.Derrida's
justification
of his framing of the "Lacan" he is readingas neither
beinglimitedto the "Seminar"noras including Lacan'slaterwork,
itselfobeysthecontradictory logicof the frame:on theone hand,
DerridawillstudythatpartofLacan'sworkwhichseemsto embody
a systemoftrutheventhoughotherwritings mightputthatsystem
in question,and on the otherhandthissamepartofLacan'swork,
says Derrida,will probablysome day be called the workof the
"youngLacan" by "university typeseagerto divideup whatcannot
be divided."WhateverDerridaactuallythinkshe is doinghere,his
contradictorywayofexplaining it obeystheparadoxesofparergonal
logic so perfectly that this self-subversion may have even been
deliberate.
If the questionof the framethus problematizes the object of
any interpretationby settingit at an angleor fold[pli] withitself,
thenDerrida'sanalysiserrsnot in opposingthisparadoxicalfuinc-
tioningto Lacan's allegoricalreading,but in notfollowing thecon-
sequencesof its own insightfarenough.If the frameis thatwhich
makesit impossibleforus to knowwhereto beginand whento
stop,forexample,whydoes Derridastop withinthe limitsof the
Dupin trilogy?And if the purposeof studying "writing" is to sow
an uncannyuncertainty about our positionin the abyss,isn'tthe
disseminallibraryDerridadescribesstill in a way just a bit too
comfortable?
"The PurloinedLetter,"says Derrida,is signed "literature."
What does this mean,if not that the letter'scontents-theonly
ones we are allowedto see-are in anothertext?That thelocus of
theletter'smeaningis not in the letter,but somewhere else? That
thecontextofthatmeaningis precisely thewayin whichitscontext
is lacking,boththroughthe explicitdesignation of a properorigin
(Crebillon'sAtree)outsidethetextand through a substitutivestruc-
turefromletterto letter,fromtextto text,and frombrotherto
485
Yale French Studies
broather,within the text, such that the expressions "outside" and

"within" have ceased to be clearly definable? But until we have
actually opened that other text, we cannot know the modalityof
the precise othernessof the abyss to itself,the way in which the
story'sedges do not simplycrumbleaway.
In order to escape the reductionof the "library"to its thematic
presenceas a sign of writing,let us thereforepull some of the books
offthe shelves and see what they contain. This is a track neither
Lacan nor Derrida has taken, but we will soon see how it in some
way enfoldsthem both.
First of all, the name "Dupin" itself,accordingto Poe scholars,
comes out of Poe's interiorlibrary: fromthe pages of a volume
called Sketches of Conspicuous Living Charactersof France (Phila-
delphia: Lea & Blanchard,1841), which Poe reviewedforGraham's
Magazine during the same month his firstDupin story appeared.
Andre-Marie-Jean-Jacques Dupin, a minor French statesman, is
there described as preciselyhimselfa walking library: "To judge
from his writings,Dupin must be a perfect living encyclopedia.
From Homer to Rousseau, fromthe Bible to the civil code, from
the laws of the twelve tables to the Koran, he has read everything,
retainedeverything..." (p. 224). Detective Dupin's "origin"is thus
multiplybookish: he is a reader whose writerread his name in a
book describinga writeras a reader-a reader whose nature can
only be describedin writing,in fact,as irreduciblydouble: "He is
the personageforwhom the paintersof politicalportraits,make the
most enormousconsumptionof antithesis.In the same picture,he
will be drawn as both great and little,courageous and timid,trivial
and dignified,disinterestedand mercenary,restive and pliable,
obstinateand fickle,whiteand black; thereis no understandingit"
(p. 210). And the writingwhichserves as the vehicleof this descrip-
tion of writtendescriptionsof double Dupin, is itself double: a
translation,by a Mr. Walsh, of a series of articlesby a Frenchman
whose name is not even knownto the translator,but who is said to
486
BarbaraJohnson
call himself"an hommede rien,a nobody"(p. 2). "Nobody"thus

becomesthepropernameof theoriginalauthorin the series.15
But the authorof the last wordin "The PurloinedLetter"is
clearlynot nobody.It is not even Poe; it is Cretbillon. What is
remarkable aboutCrebillon'sAtre'e,whenreadas thecontextfrom
whichDupin's letterto the Ministerhas been purloined,is not
simplythatit tellsthestoryof revengeas a symmetrical repetition
oftheoriginalcrime,butthatit does so precisely by meansof... a
purloinedletter.It is a letterwhichinformsKing Atreusof the
extentof his betrayal, and servesas an instrument of his revenge;
it is the Kinghimselfwho has purloinedtheletter-written by the
Queento herloverThyestesjustbeforeherdeath.The letterreveals
thatPlisthenes, whomeveryone believesto be Atreus'son,is really
theson ofhis brotherThyestes.Havingkepttheletterand its mes-
sage secret for twentyyears,Atreus plans to forcePlisthenes,
unawareofhistrueparentage, to commitpatricide. Thwartedin this
plan by Plisthenes'refusalto kill the fatherof his beloved,
Theodamia,who is, unknownto him,his sister,Atreusis forced
to producethe letter,reunitethe illicitfamily,and transfer his
revengefromPlisthenes'patricideto Thyestes'infantophagy. A
Queen betraying a King,a letterrepresenting thatbetrayalbeing
purloined forpurposesofpower,an eventualreturn ofthatletterto
itsaddressee,accompanied byan act ofrevengewhichduplicates the
originalcrime-"The PurloinedLetter"as a storyof repetition is
itselfa repetitionof thestoryfromwhichit purloinsitslast words.
The Freudian"truth"of the repetition compulsionis not simply
illustrated in the story; it is illustratedby the story.The story
obeystheverylaw it conveys;it is framedby itsowncontent.And
thus "The PurloinedLetter"no longersimplyrepeatsits own
"primalscene": whatit repeatsis nothing less thana previousstory
of repetition. The "last word" names the place wherethe non-
firstness of the "firstword"repeatsitself.
15 In a finaltwistto this mise en abyme of writing,the words "by L. L.

de Lomdnie" have been penciled into the Yale library'scopy of this book
under the title in a meticulous nineteenth-century hand, as the book's
"supplementd'orgine"...
487
Yale FrenchStudies
This is not the onlyinstanceof the folding-in of the frameof

references upon the purloinedletter'sinterior.Anotherallusion,
somewhatmore hidden,is containedin the descriptionof the
Ministeras someone"who dares all things,thoseunbecoming as
well as thosebecominga man" (Poe, p. 201). These wordsecho
Macbeth'sprotestation to his ambitiouswife: "I dare do all that
may becomea man./Who dares do more is none" (I, vii). The
reference to Macbethsubstantiates Lacan's readingof the descrip-
tionoftheMinisteras pointing towardfemininity: it is indeedLady
Macbethwho daresto do whatis unbecoming a man.And whatis
Lady Macbeth doing when we firstcatch sightof her? She is
readinga letter.Not a purloinedletter,perhaps,but one which
containsthe ambiguousletterof destiny,committing Macbethto
themurderoftheKing,whoseplace Macbethwilltake,and whose
fatehe will inevitably share.The King seemsto be preciselythat
whichcannotremainintactin thefaceof a letter:Atreusbetrayed
by his wife'sletterto his brother;Duncanbetrayedby Macbeth's
letterto Lady Macbeth; Macbethhimselfbetrayedby his own
confidence in hisabilityto readtheletterofhisFate.Andofcourse,
the King in the "PurloinedLetter,"whose power is betrayed
preciselyby his not evenknowingaboutthe existenceof the letter
thatbetrayshim.
The questionsraisedbyall thesetextstogether are legion.What
is a man?Who is the child'sfather?Whatis the relationbetween
incest,murder, and thedeathofa child?Whatis a king?How can
we read the letterof our Destiny? What is seeing? ... The cross-
roadswherethesestoriescometogether seemsto pointto thestory
of what occurredat anothercrossroads:the tragedyof Oedipus
Rex. We seemto have returned to our starting
point,then,except
forone thing:it is no longer"The PurloinedLetter"whichrepeats
thestoryofOedipus,but thestoryofOedipuswhichrepeatsall the
letterspurloinedfrom"The PurloinedLetter"'s abyssalinterior.
But that is not wherethe letterstops.For the veryOedipal
reading which Derrida attributesto Lacan is itself,according to
Derrida,a purloinedletter-purloined
by Lacan fromMarie Bona-
488
BarbaraJohnson
parte'spsychobiographical studyofthelifeand worksofEdgarAllan

Poe: "At themomentwhentheSeminar, likeDupin,findstheletter
whereit is to be found,betweenthe legs of the woman,the deci-
pheringof the enigmais anchoredin truth(...) Whythendoes it
find,at thesametimethatit findstruth,thesamemeaningand the
same toposas Bonapartewhen,leapingoverthe text,she proposes
a psycho-biographical analysisof 'The PurloinedLetter' ?" (PT,
p. 66). In that analysis,Bonapartesees Dupin's restitutionof the
letterto the Queen as the returnof the missingmaternalpenisto
the mother.The letter'shidingplace in the Minister'sapartment,
moreover, is "almostan anatomicalchart"of the femalebody-
whichleadsBonaparteto notethatBaudelaire'stranslation of"hung
froma littlebrassknobjustbeneaththemiddleofthemantelpiece"
as "suspendua un petitboutonde cuivreau dessusdu manteaude
la cheminee"["above the mantelpiece"]is "completelywrong"
(quotedin PT, p. 68). Bonaparte'sframeof reference-thefemale
body-cannot toleratethis errorof translation.
It is by meansof a notewhichLacan dropson thesubjectofthe
letter'spositionthatDerridais hereable to frameLacan forneg-
lectingto mentionhis references:"The questionof deciding," says
Lacan, "whetherhe [Dupin]seizes it [theletter]above themantel-
pieceas Baudelairetranslates,or beneathit, as in theoriginaltext,
maybe abandonedwithoutharmto the inferences of thosewhose
professionis grilling[aux inferencesde la cuisine]." * *Lacan's note:
"And even to the cook herself"(SPL, pp. 66-67).In this cavalier
treatment of Bonaparteas the "cook," Lacan thus "makesclear"
to Derrida"thatLacan had read Bonaparte,althoughthe Seminar
never alludes to her. As an authorso carefulabout debts and
he couldhaveacknowledged
priorities, thatorientshis
an irruption
entireinterpretation, as the
namelythe processof rephallization
propercourseof theletter,the 'returnof the letter'restoredto its
afterhavingbeen foundbetweenthelegs of the man-
'destination'
telpiece"(PT, p. 68). The interpretation
oftheletter(as thephallus
whichmustbe returnedto the mother)mustitselfbe returned
to the "mother"fromwhomit has been purloined-MarieBona-
489
Yale FrenchStudies
parte.Derridahimselfthusfollowspreciselythe logic he objects

to in Lacan, the logic of rectificationand correction:"to return
theletterto itspropercourse,supposing thatits trajectoryis a line,
is to correcta deviation,
to rectify
a divergence,to recalla direction,
an authenticline" (PT, p. 65). But the merefact that Derrida's
critiquerepeatsthe same logiche denouncesis in itselfless inter-
estingthanthefactthatthisrectification itselfpresupposes another,
whichputsits veryfoundations in question.For whenLacan says
thatthe questionof the exactpositionof theletter"maybe aban-
doned withoutharm"to the grillers,Derridaprotests,"Without
harm?On the contrary, the harmwould be decisive,withinthe
Seminaritself:on the mantelpiece, thelettercould not have been
'betweenthe cheeksof the fireplace', 'betweenthe legs of the fire-
place'" (PT, p. 69). Derrida mustthuscorrectLacan's text,eliminate
its apparentcontradiction, in orderto returntheletterof theinter-
pretationto its rightful owner.And all this in orderto criticize
Lacan's enterprise as one of rectification and circularreturn.If
"rectification" as such is to be criticized,it is thus difficult
to
determine whereit beginsand whereit ends. In rectifying Lacan's
textin orderto makeit fitintothe logic of rectification, Derrida
thusproblematizes theverystatusof theobjectof his criticism.
But if the correction of Lacan's textis thusitselfa mutilation
whichrequirescorrection, howare we to interpret thecontradiction
betweenLacan's descriptionof the Minister'sapartmentas "an
immensefemalebody" (SPL, p. 66) and his statementthat the
letter'sexact locationdoes not matter?This, it seems to me, is
thecruxofthedivergence betweenDerrida'sand Lacan'sinterpreta-
tionof whattheequation"letter= phallus"means.
For Bonaparte, it was preciselytheanalogybetweenthefireplace
and thefemalebodywhichled to theletter'sphallicfunction. The
phalluswas consideredas a real, anatomicalreferent servingas
the modelfora figurative representation.Bonaparte'sframeof ref-
erencewas thusreference itself.
For Derrida,on the otherhand,the phallus'frameof reference
is precisely"psychoanalytical theory"'s wayof preservingthephal-
490
Barbara Johnson
lus' referentialstatus in the act of negatingit. In commentingon

Lacan's discussionof "The Meaningof the Phallus," Derrida writes:
Phallogocentrism is one thing.And what is called man and what is called

woman mightbe subject to it. The more so, we are reminded,since the
phallus is neithera phantasy("imaginaryeffect")nor an object ("partial,
internal,good, bad"), even less the organ,penis or clitoris,which it sym-
bolizes [6crits, p. 690]. Androcentrismought thereforeto be something
else.
Yet what is going on? The entirephallogocentrism is articulatedfrom
the starting-point of a determinatesituation(let us give this word its full
impact) in which the phallus is the mother'sdesire inasmuchas she does
not have it. An (individual,perceptual,local, cultural,historical,etc.) situa-
tion on the basis of whichis developed somethingcalled a "sexual theory":
in it the phallus is not the organ, penis or clitoris,which it symbolizes;
but it does to a larger extentand in the firstplace symbolizethe penis.
(...) This consequence had to be traced in order to recognizethe meaning
(the direction,sens] of the purloinedletterin the "course which is proper
to it" (PT, pp. 98-99.)
Thus, says Derrida, it is the very non-referentiality of the phallus

which,in the finalanalysis,insuresthat the penis is its referent.
Before tryingto determinethe applicabilityof this summaryto
Lacan's actual statementsin "The Meaning of the Phallus"-not to
mentionin the "Seminar"-let us follow its consequences further
in Derrida's critique. From the very firstwords of "The Purveyor
of Truth," psychoanalysisis implicitlybeing criticized for being
capable of findingonly itself whereverit looks: "Psychoanalysis,
supposing,finds itself" (PT, p. 31, translationmine). In whatever
it turnsits attentionto, psychoanalysisseems to recognizenothing
but its own (Oedipal) schemes. Dupin findsthe letterbecause "he
knows that the letterfinallyfinds itselfwhere it must be found in
order to returncircularlyand adequately to its proper place. This
properplace, known to Dupin and to the psychoanalystwho inter-
mittentlytakes his place, is the place of castration" (PT, p. 60;
translationmodified).The psychoanalyst'sact, then,is one of mere
recognitionof the expected, a recognitionwhich Derrida finds
explicitlystated as such by Lacan in the underlinedwords he quotes
fromthe Seminar: "Just so does the purloinedletter,like an im-
491
Yale FrenchStudies
mensefemalebody,stretchout across the Minister'sofficewhen

Dupinenters.Butjustso does he alreadyexpectto findit [emphasis
mine-J.D.] and has only,withhis eyesveiledby greenlenses,to
undressthat huge body" (PT, pp. 61-62; emphasisand brackets
restored).
But if recognition is a formof blindness,a formof violenceto
the othernessof the object,it would seem that,by eliminating
Lacan'ssuggestion of a possiblecomplication of thephallicscheme,
and by lyingin waitbetweenthebracketsof the fireplace to catch
the psychoanalyst at his own game,Derrida,too, is "recognizing"
ratherthanreading.And whathe recognizes is, as he himself
states
it,a certainclassicalconception ofpsychoanalysis: "Fromthebegin-
ning,"writesDerridaearlyin his study,"we recognizetheclassical
landscapeof appliedpsychoanalysis" (PT, p. 45; emphasismine).It
wouldseem thatthe theoretical frameof reference whichgoverns
recognition is a constitutiveelementin the blindnessof any inter-
pretative insight.And it is preciselythatframeof reference which
allowsthe analystto frametheauthorofthetexthe is readingfor
practiceswhose locus is simultaneously beyondthe letterof the
textand behindthe visionof its reader.The readeris framedby
his own frame,but he is not even in possessionof his own guilt,
sinceit is thatwhichprevents his visionfromcoinciding withitself.
Justas the authorof a criminalframetransfers guiltfromhimself
to anotherby leavingsignswhichhe hopeswill be read as insuf-
ficientlyerasedtracesor referents leftby the other,the authorof
any critiqueis himselfframedby his own frameof the other,no
matterhow guiltyor innocenttheothermaybe.
Whatis at stakehereis thusthequestionoftherelationbetween
referentialityand interpretation. And here we findan interesting
twist: whilecriticizing Lacan's notionof the phallusas beingtoo
referential,Derridagoes on to use referential logicagainstit. This
comesup in connection withtheletter'sfamous"materiality" which
Derridafindsso odd. "It wouldbe hardto exaggerate herethescope
of thisproposition of the letter,or ratheron
on the indivisibility
its identityto itself inaccessible to dismemberment(...) as well
492
BarbaraJohnson
as on theso-calledmateriality of thesignifier(theletter)intolerant
But wheredoes thisidea comefrom?A torn-up
to partition. letter
maybe purelyand simplydestroyed, it happens..." (PT, pp. 86-87;
translationmodified).The so-calledmateriality of thesignifier,says
Derrida,is nothingbut an idealization.
But whatif the signifier were preciselywhatputs the polarity
in question?Has it not becomeobviousthat
"materiality/ideality"
neitherLacan's description("Tear a letterinto little pieces, it
remainstheletterthatit.is") norDerrida'sdescription ("A torn-up
lettermaybe purelyand simplydestroyed, it happens...") can be
read literally?Somehow,a rhetorical fold[plh]in the textis there
to tripus up whichever waywe turn.Especiallysincetheexpression
"ithappens"[paarrive]uses theverywordon whichthecontroversy
overthe letter'sarrivalat its destinationturns.
Our studyof the readingsof "The PurloinedLetter"has thus
broughtus to the pointwheretheword"letter"no longerhas any
literality.
But what is a letterwhichhas no literality?
5. A "Phi"forUnderstanding
I pull in resolution,and begin
To doubt the equivocation of the fiend
That lies like truth.
Macbeth
"Why do you lie to me sayingyou'regoingto

Cracow so I should believe you're goingto
Lemberg,when in realityyou are goingto
Cracow?"
-Joke quoted by Lacan afterFreud
The letter,then,is thatwhichposes the questionof its own

rhetorical through
status.It movesrhetorically thetwolong,minute
studiesin whichit is presumedto be theliteralobjectof analysis,
withouthavinganyliterality.Insteadof simplybeingexplainedby
those analyses,the rhetoricof the letterproblematizesthe very
modeofanalyticaldiscourseitself.And if"literal"means
rhetorical
493
Yale FrenchStudies
"to theletter,"the literalbecomesthemostproblematically figura-

tivemodeof all.
As thelocus of rhetoricaldisplacement,
in fact,thelettermade
its veryentranceintoPoe's storyby "traumatizing" the Prefect's
discourseaboutit.Aftera seriesofparadoxesand pleasforabsolute
secrecy,thePrefectdescribestheproblemcreatedby theletterwith
of periphrases
a proliferation whichthe narratordubs "thecantof
diplomacy":
"Well, then; I have received personal information,from a very high

quarter,that a certaindocumentof the last importancehas been purloined
fromthe royal apartments.The individualwho purloinedit is known; this
beyond a doubt; he was seen to take it. It is known,also, that it still
remainsin his possession."
"How is this known?" asked Dupin.
"It is clearly inferred,"replied the Prefect,"from the nature of the
document,and from the non-appearanceof certain results which would
at once arise fromits passing out of the robber's possession-that is to
say, from his employingit as he must design in the end to employ it."
"Be a little more explicit,"I said.
"Well, I may ventureso far as to say that the paper gives its holder
a certain power in a certain quarter where such power is immensely
valuable." The Prefectwas fond of the cant of diplomacy.(Poe, p. 200.)
The letterthusentersthediscourseofPoe's storyas a rhetorical

foldwhichactuallyhidesnothing, since,althoughwe neverfindout
whatwas written in theletter,it is presumable thattheQueen,the
Minister,Dupin,thePrefect-whoall heldtheletterin theirhands-
and even the narrator, who heardwhatthe Prefectread fromhis
memorandum-book, did. The way in whichthe letterdictatesa
seriesofcircumlocutions, then,resembles thewayin whichthepath
of theletterdictatesthe characters' circumvolutions-not thatthe
letter'scontentsmust remainhidden,but that the questionof
whetheror nottheyare revealedis immaterial to the displacement
thelettergoverns.The character and actionsof each of theletter's
holdersare determined by the rhetorical spotit putsthemin whe-
theror notthatspotcan be readby the subjectsit displaces.
The letter,then,acts as a signifier notbecauseits contentsare
its
lacking,butbecause function is notdependent on theknowledge
494
BarbaraJohnson
or nonknowledge of thosecontents.WhatLacan meansby saying

thatthe lettercannotbe dividedis thusnot thatthe phallusmust
remainintact,but thatthe phallus,the letter,and the signifier are
not substances.The lettercannotbe dividedbecauseit onlyfunc-
tionsas a division.It is not something with"an identityto itself
inaccessibleto dismemberment" as Derridainterprets it; it is a
It is knownonlyin itsefects.The signifier
difference. is an articula-
tionin a chain,notan identifiable unit.It cannotbe knownin itself
because it is capable of "sustaining itselfonlyin a displacement"
(SPL, p. 59; emphasismine).It is localized,but onlyas the non-
generalizablelocus of a differential relationship. Derrida,in fact,
enactsthislaw of thesignifier in the very act of opposingit:
Perhaps only one letter need be changed, maybe even less than a
letter in the expression: "missing from its place" (manque a sa place].
Perhaps we need only introduce a written"a", i.e. without accent, in
order to bring out that if the lack has its place [le manque a sa place]
in this atomistic topology of the signifier,that is, if it occupies therein
a specificplace of definitecontours,the order would remain undisturbed.
(PT, p. 45)
While thus criticizingthe hypostasisof a lack-the letteras the

substanceof an absence-(which is not what Lacan is saying),
whatLacan is sayingaboutboththematerial-
Derridais illustrating
ity and the localizabilityof the signifiera's the mark of difference
on theletteras a materiallocus of differentiation:
by operating by
removingthe little signifier"'," an accent mark which has no
meaningin itself.16
The questionof the natureof the "lack," however,bringsus
back to thecomplexities ofthemeaningand place of the"phallus."
For whileit is quite easy to show the signifier
as a "difference"
ratherthan a "lack," the questionbecomesmuchmoretrickyin
relationto the phallus.Therewould seem to be no ambiguity in
16 It is perhaps not by chance that the question here arises of whether

or not to put the accent on the letter"a." The letter "a" is perhaps the
purloinedletterpar excellencein the writingsof all three authors: Lacan's
"objet a," Derida's "differance,"and Edgar Poe's middle initial, A, taken
fromhis fosterfather,JohnAllar,
495
Yale FrenchStudies
Lacan's statement that"Clinicalobservation showsus thatthistest

through thedesireoftheOtheris notdecisiveinsofaras thesubject
therebylearnswhetheror not he himselfhas a real phallus,but
insofaras he learnsthatthemotherdoesn't"(Wtrits, p. 693; trans-
lationand emphasismine).The theoryseemsto implythatat some
pointin humansexuality,a referential momentis unbypassabie:
theobservation thatthemotherdoes nothavea penisis necessary.
And therefore it would seem thatthe "lack" is localizableas the
substanceof an absenceor a hole.To borrowa joke fromGeoffrey
Hartman'sdiscussionof certainsolutionless detectivestories,ifthe
purloinedletteris the mother'sphallus,"insteadof a whodunitwe
get a whodonut, a storywitha hole in it.""1
But even on this referential level,is the object of observation
reallya lack? Is it not rathernotan observation at all but already
("castration") not of a lack but
an interpretation-aninterpretation
preciselyofa difference? If whatis observedis irreducibly
anatom-
ical,whatis anatomyherebut theirreducibility ofdifference?
Even
on the mostelementary level,the phallusis a signof sexualityas
difference,and notas thepresenceor absenceofthisor thatorgan.
But Lacan definesthe phallusin a muchmorecomplicated way
thanthis.For if the womanis definedas "givingin a love-relation
thatwhichshe does not have,"the definition of whatthe woman
does not have is not limitedto the penis.At anotherpointin the
discussion,Lacan refersto "the giftof what one does not have"
as "love" (6crits,p. 691). Is "love" here a meresynonym of the
phallus?Perhaps.Butonlyifwe modify thedefinitionofthephallus.
"Love"is,inLacan'sterminology, whatis inquestionin the"request
forlove" (demanded'amour],whichis "unconditional," the"demand
fora presenceor an absence"(tcrits,p. 691). This demandeis not
onlya reference to "whatthe Otherdoesn'thave,"however.It is
also language.And languageis whatalienateshumandesiresuch
that"it is fromtheplace of theOtherthatthesubject'smessageis
17 GeoffreyHartman, "Literature High and Low: the Case of the

MysteryStory,"in The Fate of Reading,Universityof Chicago Press, 1975,
p. 206.
496
BarbaraJohnson
emitted"(J8crits, p. 690). The "demand"is thusa requestforthe

unconditional presenceor absencenotof an organbut of theOther
in answerto the questionasked by the subjectfromthe place of
the Other.But thisdemandeis not yet the definition of "desire."
Desireis whatis leftof thedemandewhenall possiblesatisfaction
of "real" needshas been subtracted fromit. "Desireis neitherthe
appetitefor satisfaction, nor the demandfor love, but the dif-
ferencewhichresultsfromthe subtraction of the firstfromthe
second,the very phenomenonof their split [Spaltung]"(18crits,
p. 691). And if thephallusas a signifier,
accordingto Lacan,"gives
theratioof desire,"thedefinition ofthephalluscan no longerbear
a simplerelationeitherto the body or to language,because it is
thatwhichprevents boththebodyand languagefrombeingsimple:
"The phallusis the privilegedsignifier of thatmarkwherelogos
is joined togetherwiththe adventof desire"(ecrits,p. 692; all
translations in thisparagraph mine).
The important wordin thisdefinition is "joined."For iflanguage
(alienationof needsthrough theplace of theOther)and desire(the
remainder whichis leftfromthe subtraction of the satisfactionof
realneedsfromabsolutedemand)are neithertotallyseparablefrom
each othernor relatedin the same way to theirown division,the
phallusis the signifier of the articulationbetweentwo verypro-
blematicchains.Butwhatis a signifier in thiscontext?"A signifier,"
says Lacan, "is whatrepresents a subjectforanothersignifier." A
signifier represents,then,and what it represents is a subject.But
it onlydoes so foranothersignifier. Whatdoes theexpression"for
anothersignifier" mean,ifnot thatthe distinction betweensubject
and signifier posed in the firstpartof the definition is beingsub-
vertedin the second? "Subject"and "signifier" are co-implicated
in a definitionwhichis unableeitherto separatethemtotallyor to
fuse themcompletely. Thereare threepositionsin the definition,
two of whichare occupiedby the sameword,but thatwordis dif-
ferentiated fromitselfin the courseof the definition-because it
beginsto take the place of the otherword.The signifier forwhich
the othersignifier represents a subjectthusacts like a subjectbe-
497
Yale FrenchStudies
cause it is theplace wherethe representation is "understood." The

signifier,then,situatesthe place of something like a reader.And
thereaderbecomestheplacewhererepresentation wouldbe under-
stoodiftherewereanysuchthingas a placebeyondrepresentation;
the place whererepresentation is inscribedas an infinite chainof
substitutions whetheror not thereis any place fromwhichit can
be understood.
The letteras a signifieris thusnot a thingor the absenceof a
thing,nor a wordor the absenceof a word,nor an organor the
absenceof an organ,but a knotin a structure wherewords,things
and organs can neitherbe definablyseparatednor compatibly
combined.This is whythe exact representational positionof the
letterin theMinister'sapartment bothmattersand doesnotmatter.
It mattersto the extentthatsexual anatomicaldifference creates
an irreducible dissymmetry to be accountedfor in everyhuman
subject.But it does not matterto the extentthatthe letteris not
hiddenin geometrical space, wherethe police are lookingforit,
or in anatomicalspace, wherea literalunderstanding of psycho-
analysismightlook forit. It is located"in" a symbolicstructure, a
structurewhichcan only be perceivedin its effects,and whose
effectsare perceivedas repetition. Dupinfindsthe letter"in" the
symbolicordernot because he knowswhereto look,but because
he knowswhatto repeat.Dupin's"analysis"is therepetition of the
scenewhichled to thenecessityof analysis.It is not an interpreta-
tionor an insight, but an act. An act of untying the knotin the
structure by meansof the repetition of the act of tyingit. The
word "analyze,"in fact,etymologically means"untie,"a meaning
on whichPoe plays in his prefatory remarkson the natureof
analysisas "thatmoralactivitywhichdisentangles" (Poe, p. 102).
The analystdoes not intervene by givingmeaning, but by effecting
a denouement.
But if the act of (psycho)analysis
has no identityapartfromits
statusas a repetition of thestructure
it seeksto analyze(to untie),
thenDerrida'sremarksagainstpsychoanalysis as beingalwaysal-
readymiseen abymein thetextit studiesand as beingonlycapable
498
Barbara Johnson
of findingitself,are not objections to psychoanalysisbut in fact a

profoundinsightinto its veryessence. Psychoanalysisis in factitself
the primal scene it is seeking: it is the firstoccurrenceof what
has been repeatingitselfin the patient withoutever having occur-
red. Psychoanalysisis not itselfthe interpretation of repetition; it
is the repetitionof a trauma of interpretation-called"castration"
or "parental coitus" or "the Oedipus complex" or even "sexuality"
-the traumaticdeferredinterpretation not of an event, but as an
event which never took place as such. The "primal scene" is not a
scene but an interpretativeinfelicitywhose result was to situate
the interpreter in an intolerableposition. And psychoanalysisis the
reconstructionof that interpretative infelicitynot as its interpreta-
tion, but as its firstand last act. Psychoanalysishas content only
insofaras it repeats the dis-contentof what never took place.
But, as Dupin remindsus, "there is such a thingas being too
profound.Truth is not always in a well. In fact, as regards the
more importantknowledge,I do believe that she is invariablysuper-
ficial"(Poe, p. 119). Have we not here been lookingbeyondLacan's
signifierinstead of at it? When Lacan insists on the "materiality
of the signifier"which does not "admit partition,"what is his way
of explainingit? Simplythat the word "letter" is never used with
a partitivearticle: you can have "some mail" but not "some letter":
Languagedeliversits judgmentto whomever knowshow to hear it:

through the usage of the articleas partitive particle.It is therethatthe
spirit
-if spiritbe livingmeaning -appears, no less oddly,as more
availableforquantification thanthe letter.To beginwithmeaningitself,
whichbears our saying:a speechrichwithmeaning["pleinde significa-
tion"],just as we recognizea measureof intention ["de l'intention"]in
an act, or deplorethatthereis no morelove ["plusd'amour"];or store
up hatred["de la haine"] and expenddevotion["du devouement"], and
so muchinfatuation ["tantd'infatuation"] is easilyreconciledto the fact
thattherewill alwaysbe ass ["de la cuisse"]forsale and brawling["du
rififi"]
amongmen.
But as for the letter-be it takenas typographical character,epistle,
or whatmakesa man of letters -we will say thatwhatis said is to be
understood to the letter[a' la lettre],thata letter[une lettre]awaitsyou
499
Yale FrenchStudies
at the post office,or even that you are acquainted with letters[que vous
avez des lettres]-never that there is letter[de la lettre]anywhere,what-
ever the context,even to designate overdue mail. (SPL, pp. 53-54.)
If thispassageis particularlyresistantto translation,

thatin itself
is a resultof the factthatits messageis in the "superficial"play
of the signifier.Like the largeletterson the map whichare so
obviousas to be invisible,Lacan's textualsignifier has gone un-
noticedin the searchforthe signified, "signifier."
But the questionof translation in connectionwitha message
whichis so obviousthatit goesunseenis notan accidenthere.For
in his discussionof Dupin'sstatement that"'analysis'conveys'al-
gebra' about as much as, in Latin, 'ambitus'implies'ambition',
'religio' religion,or 'homines honesti' a set of 'honorable men'"
(Poe, p. 212),Lacan asks:
Might not this parade of erudition be destined to reveal to us the

key words of our drama? 18 Is not the magician repeatinghis trick before
our eyes, without deceiving us this time about divulginghis secret, but
pressinghis wager to the point of really explainingit to us without us
seeing a thing.That would be the summitof the illusionist'sart: through
one of his fictivecreations to trulydelude us. (SPL, pp. 50-51.)
But thetrickdoes notend there.For has Lacan himself notslipped

into the paragraphon the quantificationof the lettera paradeof
"key words"forhis readingof the situation?"Full of meaning,"
"intention," "devotion,""ass for
"hatred,""love," "infatuation,"
sale," and "brawlingamongmen"-all of thesewordsoccuras the
possiblesignifiedsof "The PurloinedLetter"in the Seminar.But
if the key wordsof a readingof the storythusoccuronlyin the
mode of a play of the signifier,the differencebetween "signifier"
in Lacan's text,as well as in Poe's, has been effec-
and "signified"
18 Ambitus means "detour"; religio, "sacred bond"; homines honesti,

"decent men." Lacan expands upon these words as the "key words" of the
storyby saying: "All of this (.. .) does not implythat because the letter's
secrecy is indefensible,the betrayalof that secret would in any sense be
honorable. The honesti homines,decent people, will not get off so easily.
There is more than one religio, and it is not slated for tomorrowthat
sacred ties shall cease to rend us in two. As for ambitus: a detour, we
see, is not always inspiredby ambition"(SPL, p. 58).
500
BarbaraJohnson
tivelysubverted.What the readerfinallyreadswhenhe deciphers

the signifying surfaceof the map of his misreading is: "You have
beenfooled."And it is precisely in thisdiscussionof "beingfooled"
that Lacan, far fromexcludingthe narrator, situateshim in the
dynamicfunctioning of the text,as a readeren abyme.dupedby
Dupin'strickexplanations of his technique,a readerwho,however,
unconsciousof the nonsequiturs he is repeating,is so muchin awe
of his subjectthathis admiration blindsus to the trickyfunction-
ingof whathe so faithfully transmits.
To be fooledby a text impliesthatthe textis not constative
but performative, and thatthe readeris in factone of its effects.
The text's"truth"is whatputsthestatusofthereaderin question,
whatperforms himas its "address."Thus "truth"is not whatthe
fictionrevealsas a nudityhiddenbehinda veil.WhenDerridacalls
Lacan's statementthat"truthinhabitsfiction"an unequivocalex-
pressionor revelation of the truthof truth(PT, p. 46), he is simply
not seeingthe performative perversity of the restof the sentence
in whichthat "statement" occurs: "It is up to the readerto give
the letter(...) whathe will findas its last word: its destination.
That is, Poe's messagedecipheredand comingback fromhim,the
reader,fromthe factthat,in readingit,he is able to sayofhimself
that he is not morefeignedthan truthwhen it inhabitsfiction"
(tcrits,p. 10; translation
mine).The playbetweentruthand fiction,
readerand text,messageand feint,has becomeimpossibleto un-
ravelintoan "unequivocal"meaning.
We have thuscomeback to thequestionof theletter'sdestina-
tion and of the meaningof the enigmatic"last words"of Lacan's
Seminar."The sender,"writesLacan, "receivesfromthe receiver
his ownmessagein reverseform.Thus it is thatwhatthe'purloined
letter',nay,the 'letterin sufferance' meansis thata letteralways
arrivesat its destination"(SPL, p. 72). Whatthereversibility ofthe
directionof theletter'smovement betweensenderand receiverhas
now come to standforis preciselythe fact,underlined by Derrida
as if it werean objectionto Lacan, thatthereis no positionfrom
which the letter'smessage can be read as an object: "no neutraliza-
501
Yale French Studies
tion is possible, no generalpoint of view" (PT, p. 106). This is also

preciselythe "discovery"of psychoanalysis-thatthe analyst is in-
volved (throughtransference)in the very "object" of his analysis.
Everyonewho has held the letter-or even beheld it-including
the narrator,has ended up having the letter addressed to him as
its destination.The reader is comprehendedby the letter: there is
no place fromwhich he can stand back and observe it. Not that
the letter'smeaningis subjectiveratherthan objective,but that the
letter is precisely that which subverts the polarity subjective/ob-
jective,that whichmakes subjectivityinto somethingwhose position
in a structureis situated by the passage throughit of an object.
The letter's destination is thus whereverit is read: the place it
assigns to its reader as his own partiality.Its destinationis not a
place, decided a priori by the sender, because the receiver is the
sender, and the receiver is whoever receives the letter,including
nobody. When Derrida says that a letter can miss its destination
and be disseminated,he reads "destination"as a place which pre-
exists the letter's movement.But if, as Lacan shows, the letter's
destinationis not its literal addressee, nor even whoeverpossesses
it, but whoeveris possessed by it, then the verydisagreementover
the meaning of "reachingthe destination"is an illustrationof the
non-objective nature of that "destination." The very rhetoric of
Derrida's differentiation of his own point of view from Lacan's
enacts that law:
Thanksto castration,the phallus always stays in its place in the

transcendentaltopology we spokeof earlier.It is indivisible
and indestruc-
tible there,like the letterwhichtakes its place. And that is why the
interestedpresupposition,neverproved,of the letter'smateriality as in-
was indispensable
divisibility to this restricted economy,this circulation
of propriety.
The differenceI am interested
in hereis that,a formulato be read
howeverone wishes,the lack has no place of its own in dissemination.
(PT, p. 63; translation
modified,
emphasismine.)
The play of interestin this expressionof differenceis too in-

terestingnot to be deliberate.The oppositionbetweenthe "phallus"
502
BarbaraJohnson
and "dissemination" is notbetweentwo theoretical objectsbut be-

tween two interestedpositions.And if senderand receiverare
merelythe two poles of a reversiblemessage,thenLacan's very
substitution of "destin"for "dessein"in the Crebillonquotation
-a misquotation whichDerridafindsrevealingenoughto end his
analysisupon-is in factthequotation'smessage.The sender("des-
sein") and the receiver("destin")of the violencewhichpasses be-
tweenAtreusand Thyestesare equallysubjectto the violencethe
letteris.
The reflexivitybetweenreceiverand senderis, however,not an
expression ofsymmetry in itself,but onlyan evocationof theinter-
dependenceof the two terms,of the questionof symmetry as a
problemin the transferential structure of all reading.As soon as
accidentor exteriority or time or repetition entersinto that re-
flexivity-that is to say,fromthe beginning-,Othernessitselfbe-
comesin a waythe letter'ssender.The messageI am readingmay
be eithermy own (narcissistic) messagebackwardsor the way in
whichthatmessageis alwaysalreadytraversedby its own other-
ness to itselfor by the narcissistic messageof the other.In any
case the letteris in a way the materialization of my death.And
once thesevariouspossibilities are granted,noneof themcan func-
tionin isolation.The questionof theletter'soriginand destination
can no longerbe asked as such. And whetherthis is because it
involves2, 3 or 4 termsmustremainundecidable.
The sentence"a letteralwaysarrivesat its destination" can thus
eitherbe simplypleonasticor variouslyparadoxical:it can mean
"the only messageI can read is the one I send," "whereverthe
letteris, is its destination," "when a letteris read, it reads the
reader,""the repressedalwaysreturns," "I existonlyas a reader
of the other,""the letterhas no destination," and "we all die." It
is notanyone of thesereadings, but all ofthemand othersin their
veryincompatibility, whichrepeatthe letterin its way of reading
theact ofreading.Far fromgivingus theSeminar'sfinaltruth, these
last wordsenact the impossibility of any ultimateanalyticalme-
talanguage.
503
Yale FrenchStudies
If it at firstseemedpossibleto say thatDerridawas opposing

theunsystematizable to the systematized, "chance"to psychoanaly-
tical "determinism," or the "undecidable"to the "destination," the
positionsof these oppositionsseem now to be reversed:Lacan's
apparentlyunequivocalendingsays only its own dissemination,
while"dissemination" has erecteditselfintoa kindof "last word."
But these oppositionsare themselvesmisreadings of the dynamic
functioning of what is at stake here.For if the letter is precisely
thatwhichdictatestherhetorical indetermination ofanytheoretical
discourseabout it, thenthe oscillationbetweenunequivocalstate-
mentsof undecidability and ambiguousassertionsof decidability is
preciselyone of the letter'sinevitableeffects.Thus it is, for
example,thatthe "indestructibility of desire,"whichcouldbe con-
sidereda psychoanalytical beliefin the returnof the same,turns
out to name repetition as the repetitionnot of samenessbut of
otherness, whichresultsin the dissemination of the subject.And
"symbolic determination" is notopposedto "chance": it is precisely
whatemergesas thesyntaxofchance.19But "chance,"out ofwhich
whatrepeatssprings, cannotin anywaybe "known,"since"know-
ing" is preciselyone of its effects.We can therefore neverbe sure
whetheror not "chanceitself"existsat all. "Undecidability" can
no morebe used as a last word than "destination.""Car," said
Mallarm6,"il y a et ii n'ya pas de hasard."The "undeterminable"
is notopposedto thedeterminable; "dissemination" is notopposed
to repetition. If we could be sure of the difference betweenthe
determinable and the undeterminable, the undeterminable would
be comprehended withinthe determinable. What is undecidableis
preciselywhethera thingis decidableor not.
19 This is what the mathematicalmodel in the Seminar's"Introduction"

clearly shows: beginningwith a totally arbitrarybinary series, a syntax
of regularityemergesfromthe simple applicationof a law of combination
to the series. When it is objected that that syntaxis not, unless the subject
remembersthe series, Lacan responds,"That is just what is in question
here: it is less out of anythingreal (.. .), than preciselyout of what never
was, that what repeats itself springs" (tcrits, p. 43; translationmine).
Memorycould thus be considerednot as a conditionof repetition,but as
one of its syntacticeffects.What we call a randomseries is, in fact,already
an interpretation, not a given: it is not a materializationof chance itself,
but onlyof somethingwhichobeys our conceptionof the laws of probability.
504
BarbaraJohnson
As a finalfoldin theletter'sperformanceof its reader,it should

perhapsbe noted that,in this discussionof the letteras what
preventsme fromknowingwhetherLacan and Derridaare really
sayingthe same thingor onlyenactingtheirown differences from
themselves, my own theoretical"frameof reference" is precisely,
to a verylargeextent,thewritings ofLacanand Derrida.The frame
is thusframed againbypartofitscontent;thesenderagainreceives
his own messagebackwardsfromthe receiver.And thetrueother-
nessof thepurloinedletterofliterature has perhapsstillin no way
been accountedfor.
505

The Frame of Reference, Poe, Lacan Derrida - Barbara Johnson

Uploaded by

Copyright:

Available Formats

You might also like

The Frame of Reference, Poe, Lacan Derrida - Barbara Johnson

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Frame of Reference, Poe, Lacan Derrida - Barbara Johnson

Uploaded by

Copyright:

Available Formats

The Frame of Reference: Poe, Lacan, Derrida

Author(s): Barbara Johnson

The Frame of Reference: Poe, Lacan, Derrida

1. The Purloined Preface

A literarytext which both analyzes itselfand shows that it ac-

The subversionof any possibilityof a positionof analytical

reading-alsodemonstrated by each of thetexts-is smallcomfort,

In 1845,EdgarA. Poe publishedthethirdof his threedetective

"I have no betteradvice to give you," said Dupin. "You have, of

Whatis paraphrasedis thusthe description of the letterthe story

theletter'smessageis neverrevealed,whichwillserveas thebasis

There are two scenes, the firstof which we shall straightwaydesignate

4 Such a concatenation could jokingly be called, after the nursery

6 "So infamous a scheme/ If not worthy of Atreus, is worthy of

Thus,it is neitherthe characterof the individualsubjects,nor

2) What Lacan leaves out of the text. This objection is itself

L'ascendant que le ministretire de la situa-

We have just seen how Derrida,in his effortto right(write)

-Un dessein si funeste,

Atreus,whosewifewas longago seducedby Thyestes,is aboutto

places Dupin as one of the "ostriches"in the "triad."Not content

8 J. Derrida, Positions, Minuit, 1972, pp. 112-113 (translationand em-

This type of oscillation between two terms, considered as

11 Cf. Lacan's description of the "effect of disorientation, or even of

a sameness,a repetitionof the errorwhichthe encounteris designed

the necessityof substitution in the objectof his desire(the object

simulacrum to simulacrum,fromdoubleto double"(omittedin PT;

1) Whatis therelationbetweena dividedunityand a duality?

thingthat was never reallypresent).In Lacan's articleon "La

"It would appear,"commentsDerrida,"thatthe Hegelianmove-

12 Some idea of the possibilitiesfor misunderstanding inherentin this

pretationon Derrida'spart,and quite an astuteone at that,with

uncertainterms: "Thus it is thatwhatthe 'purloinedletter',nay,

Is thatall, and shall we believe we have decipheredDupin's real strategy

The meaningof thislast sentenceis problematized notso much

subvert, thepresupposition of univocality whichis theveryfounda-

13 Paul de Man, Blindness and Insight,Oxford UniversityPress, 1971,

ysis takesplace,and whichis, in fact,suggestedby the veryfirst

Theythankhimforthegrandtruths he has justproclaimed,

The factthatthisquotationfromBaudelairerefersto Poe and not

4. The Frameof Reference

If Derrida is thus framingLacan foran interpretative

frameteachus aboutthenatureoftheact ofreading, in thecontext

Withoutbreathinga word about it, Lacan excludes the textual fiction

In framingwith such violence,in cuttinga fourthside out of the nar-

It would seem,then,thatLacan is here guiltyof severalsins

By overlookingthe narrator'sposition, the narrator'sinvolvementin

14 StanleyE. Fish, "How Ordinaryis OrdinaryLanguage?,"New Literary

narrator,in narratinghimself,advances a few propositionswhich carry

Here theargument seemsto reversethepreviousobjection:Lacan

1) The textualsignifier's resistanceto beingtotallytransformed

quotationmarksthat surroundthe entirestory,Dupin is obliged

by partof its contents, it is preciselythisslippagebetweensignifier

broather,within the text, such that the expressions "outside" and

call himself"an hommede rien,a nobody"(p. 2). "Nobody"thus

15 In a finaltwistto this mise en abyme of writing,the words "by L. L.

This is not the onlyinstanceof the folding-in of the frameof

parte'spsychobiographical studyofthelifeand worksofEdgarAllan

parte.Derridahimselfthusfollowspreciselythe logic he objects