First Person, Second Person, Same Person: Narrative as Epistemology

Author(s): Mieke Bal

Reviewed work(s):
Source: New Literary History, Vol. 24, No. 2, Reconsiderations (Spring, 1993), pp. 293-320
Published by: The Johns Hopkins University Press
Stable URL: http://www.jstor.org/stable/469408 .
Accessed: 03/12/2012 13:13

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 [email protected].

The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and extend access to
New Literary History.

http://www.jstor.org

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
 FirstPerson, Second Person, Same Person:
NarrativeAs Epistemology
Mieke Bal

I. Introduction

L ET ME START with two statementsabout description. "Tran-

scription always also description,"writesanthropologistJo-
is
hannes Fabian in his seminal Power and Performance;'and
continuingthis idea, philosopher and historianof art Hubert Dam-
isch writes: "Describing, in this sense, is always already narrating"
in his equally seminal L'originede la perspective.2
Something seems to
be the matter,epistemologicallyspeaking, with description.Joining
the one statementto the other, something seems to be a contami-
nation that infects"pure" neutral rendering with the taint of nar-
rativity.Fabian takes descriptionas already implicatedin thisprocess,
but warns that even transcriptionis "always part of a process of
interpretationand translation"(P 110).
What else can a historical,philosophical,erudite treatiseon linear
perspective,and its origin in Italian art (history),have in common
with a critical-anthropologicalstudy of a proverb and a theatrical
performancebased on it,in Shaba, Zaire? What could make it worth
analyzing the common elements in two so diverse scholarly texts,
neither of which is on narrativenor is a narrativein any common-
sense definition?Both are products of academic work, of "new art
history"and of "critical anthropology" respectively;both, that is,
stand in the tradition of contemporary,"progressive" knowledge
production.
Let me brieflytell these two stories.One evening at dinner,Fabian
heard the expression, probably proverbial,le pouvoirse mangeentier
(power is eaten whole). Trying to find out what the proverb means
in the Shaba culture,he asked people, and one day (second episode)
he asked a group of theater actors who were his friends.Afteran
intense session of brainstorming,the group decided to work up
theirnext play around the saying.Fabian was presentat preparations,
rehearsals,and performance,consideringthat what happened is the
best possible form of modern ethnography: the constructionof

New Literary
History,1993, 24: 293-320

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
294 NEW LITERARY HISTORY

knowledge about a culture withthe people and through collective

research and discovery.The real performanceof the actorsbecomes
an allegory of the idea of performanceas an epistemic model for
ethnography.
Damisch's pursuitof knowledgeconcernsthe originof perspective
as well as of the thinkingabout perspective-perspective as a dis-
course-and proceeds on an equally "democratic"base. He provides
close readings of the treatiseson perspectives,the experimentsthat
led to what can only be anachronisticallycalled its "discovery,"and
of three paintings, constructed as a group, which he studies in
relation to one another in order to understand the origin of per-
spective through the transformationsin its use.
Both of these books address up front the basic epistemological
problem of their discipline and, I contend, of the humanitiesand
the social sciences in general. For Fabian, that problem is to account
not only for knowledge as a product but for its production in an
epistemicsituationwhere power inequalityhas made the discipline's
traditionalparadigms virtuallyuseless. For Damisch, the problem is
the paradox of historicalsearch foran origin.To avoid mythification,
origin must be seen through a double predicament. It presupposes
a beginning which must be revolutionaryin order to be perceived,
yet must be absorbed in a traditionin order to measure its revo-
lutionary impact; an absorption which requires that tradition to
ignore the event which it acknowledges as its origin (L 79).
The rationale for my choice of discussing these texts is the
relevance to them of, and their relevance for, a third text: a phil-
osophical analysisof epistemologicalproblemsin relationto feminist
theory,Lorraine Code's 1991 WhatCan She Know?3That relevance
is best seen when one realizes that both case studies grew out of a
search for an epistemologythrough which the subjective status of
the objects the writerssought to understand could be done more
justice. Both books grew out of an impatience, as Damisch put it
in his firstsentence,withthe epistemologicalmodes currentin their
respectiveareas, anthropologyand history;an impatience,also, which
gives the books an autobiographical slant.
Johannes Fabian is a well-knowncriticalanthropologist:his Time
and the Other4is perhaps the single most important text of the
movement of anthropologistsimpatient with and thoroughlysus-
picious of the colonialistlegacy that subtends their field. His recent
Powerand Performance, the textunder discussion in this paper, gives
the lie to those who become skeptical of what they consider the
overcriticalmood of the relentlesscritiqueof ethnography;to those
who tend to conclude from Fabian's criticalwork that there is no

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
NARRATIVE AS EPISTEMOLOGY 295

way one can do it right. Here, the same author who would never
be satisfiedwith the well-meaningattemptsat democraticand non-
exploitativeethnographies is showing his hand. And whatever else
this seminal study may also be, it remains true to the discursive
habitsof ethnography:it is a narrativetext.A complex, multilayered
and intricate one indeed, telling the multiple stories of his own
discoveries of ethnographic factsas well as of methodology,mixed
withthose of the discussionsand rehearsals,interlardedwithstories
about each of the participantsin the project as well as of popular
theater in Shaba. Complex, yes, but a narrativeno less. And since
it offers an in-depth exploration of ways of pursuing knowledge,
the embedding part of the narrativeis self-reflexive.
In addition to being a philosopher, Damisch is primarilyan art
historian,and as such a teller of tales too. His masterfulTheoriedu
nuage5(Theory of the Cloud) was as pathbreakingas Fabian's Time
and the Other.In that book he took clouds to be, as he formulates
it in his later study on perspective,"emblems of what perspective
excludes from its order . . . while also of the logic on which it is
based and which gives it coherence" (L 297). Moreover, his topic is
the search for an origin, and states as much in its title, and that
topic promises a text that is at least doubly narrative: as storyof
origin and as story of the search for that origin. His impatience
also concerned his fellow historians and their simplisticconcep-
tion of what kind of narrativeshistoryought to construct."Hence,
this study, too, presents a complex narrative with a self-reflexive
dimension.
Both studiesrepresentan object lyingrigorouslyoutside the subject
of inquiry: a discursive habit of a differentpeople in terms of
cultural identityand location, the people of Shaba in Zaire, in the
one case, a differentdiscursive apparatus in terms of time, in
renaissance Italy, in the other. Both are explicitlyengaged in over-
coming the object status this "third personhood" entails. Fabian's
intention"to explore the meanings of 'le pouvoir se mange entier'
and to do this followinga 'method' that works as an ethnography
with, not of, the Groupe Mufwankolo" (P 55) echoes Damisch's
search for "an analysis which would be less about a painting than
it would have to reckon withit" (L 240). And there, too, lies their
relevance for the kind of inquiryLorraine Code proposes, into the
relationshipsbetween subject and object of knowledge whose pos-
sibilityemerges when the traditional objectivism'sself-evidenceis
suspended.
Thus at firstsight, in terms of narratologyand seen from the
perspective of the object of study,both narrativescan be seen as

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
296 NEW LITERARY HISTORY

primarily"third-personnarratives,"including the tensionsinherent

in that discursive mode. But both authors struggle with precisely
that dimension of their professionaldiscourses: the false neutrality
sufficientlychallenged by contemporaryepistemologyas it translates
into a narrativetold in the third person, with an invisiblenarrator
and a nonidentifiedfocalizer. First, they turn their texts around,
and the self-reflexiveside of the studies involves the firstperson
in their quest. The neutral, distant narrator becomes part of the
exploration, so that the embedding narrativeis writtenin the first
person.
Where Damisch opens his preface with the statement of his
impatience, Fabian espouses the conventionsof realistfictionwhen
he opens his firstchapter thus: "On the evening of June 17, 1986,
in the midst of a relativelyshortstintof field work in Lubumbashi,
the capital of the mining region of Shaba in Zaire, I was writing
up the day's events when I made a discovery" (P 3). Placing the
events to follow in a specific time frame, himself as the story's
narrator in specificcircumstances,and the firstevent of the fabula
to come as the potentiallyspectacular interruptionof a durative
occupation: it could be practicallyany novel. His identityas a self-
reflexivenarrator represented as writingspecifiesthe discourse, as
it suggests a modernistaesthetic.
But between a realist, neutral narrator and a modernist first
person, the problem of the statusof the object of narrationremains
as yet to be examined. In the case of epistemological narratives,
the issue becomes that of the subjective status of the object of
inquiry. Both authors make that status an importantelement in
their experimental narrative. Fabian sets up a situation in which
the object, cultural knowledge, is not studied but constructed on
the spot, by and with and through the cultural group under in-
vestigation,as well as conductingthat investigation.This is precisely
the discoveryalluded to: that ethnographicknowledge is not simply
a dialogue, let alone a neat and clean third-personnarrative,but
a performance. Damisch, whose object is not a cultural group but
a cultural discourse, sets up a specificenunciation of perspectiveas
his interlocutor:the group of three anonymous, renaissance, "ur-
binate," perspectival city views, the most famous of which is The
Ideal Cityat Urbino. Both scholars take the second person as the
core of theirexamination,and both make action,process,performance
the core of their knowledge. Both performformal experimentsto
inscribe this second personhood of their object as well as the per-
formativedynamic into the narrative,which therebybecomes ex-
perimental, complex, and theoreticallyat least, a second-person

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
NARRATIVE AS EPISTEMOLOGY 297

narrative. Both books, then, are semantically"third-person,"syn-

and presentattemptsto achieve pragmatically
tactically"first-person,"
a "second-person" narrative.
This intricationof a project of criticalepistemology,a narrative
whichinscribesthe second person,and the centralityof performance,
struck me as significant,and especially so in light of recent devel-
opments in epistemological theory as they relate to narrative. To
sum these up too briefly:the epistemological notion of objective
truthand impersonal knowledgeis bound up withthe narratological
notions of "third-personnarrative,"external and invisiblenarrator,
and neutral representation.But if we realize thatthe Cartesian cogito
which sustains the objective epistemologyis itselfa mininarrativein
thefirstperson, we don't even need Descartes's personal expressions
of anxietyto realize that this conception of knowledge is inherently
contradictory.'Indeed, the Cartesian principles are all bound up
with subjectivityand defined in termsof the individual subject: the
basis of knowledge is one indubitable thing to which all other
knowledge is systematicallyrelated; hence, to which it is relative;
"indubitable" presupposes a subject of possible doubt; reason is
common to and alike in all knowers; yet the quest for knowledge
is undertaken separately by each rational being who is thereby
unassisted by the senses and uses the same method. Where the
subject of inquiry is so emphaticallyand contradictorilyboth fore-
grounded and neutralized, one might well associate this epistemo-
logical ideal withwhat Philippe Lejeune analyzes as "autobiography
in the third person."8
In her fascinatinginquiry into the conditions of knowledge and
the problem of access to knowledgeforsome of those rationalbeings
who are apparently a littleless fitto be such a subject of inquiry,
Lorraine Code challenges,among many other things,the overruling
primacyof objectivityand the paradigmatic status of physicsas the
ideal model of knowledge. The two are, of course, related. The
attractionsof physics are deceptive: they consist in providing the
illusion that knowledge can always be analyzed in observational
"simples" (W 139). She proposes instead to give primacy to inter-
subjectivity("a conception of cognitive agency for which intersub-
jectivityis primaryand 'human nature' is ineluctablysocial" [W 72])
and to give paradigm status to the difficultand complex episte-
mological project of knowingotherpeople. Central in her analysis of
the knowing subject is Annette Baier's concept of secondpersonhood,
and the model for the mode of inquiry she proposes is friendship.
Since the termsofferedfor reflectionare strikinglyclose to those
put to use by Fabian and Damisch, it comes as no surprise that the

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
298 NEW LITERARY HISTORY

discursive consequencesof Code's theoretical positionseemto apply

quitespecifically to thetextsproducedbythesetwo"field"scholars.
And narrativity is the locus of theseconsequences.In the wake of
her critiqueof physicsas paradigm,Code scornfully suggeststhat
there are narrativereasons why epistemologyvalues simplicity:
"Clean, unclutteredanalysesare valued more highlythan rich,
multifacted, but messyand ambiguousnarratives"(W 169). This
remarkstrongly suggeststhatthereis a relationbetweennarrative
formand epistemological competence;betweentheabilityto handle
complexknowledge and to tell and read complexstoriesas much
as betweencleanlinessand simplicity. In otherwords,if Code is
righthere, as I think she is, then narrativetheoryand analysishave
a lot to offerin the importantarea of reflection on whatit is and
howit is we can know.And sincetheacademicendeavoras a whole
is verymuch investedin those questions,thereis an opportunity
for humaniststo contribute to the foundationsof academic,intel-
lectuallifethatI would hate to miss.
One caveatis alreadycalled for,however.Code's remarksounds
convincingnot onlybecause,on a symbolic-logical level,it is easy
to imaginehow it can be right,and on the level of indexicality,
hownonsimplistic analyseswouldrequirecomplexnarratives as their
accounts,butalsobecauseitsuggestsa resemblance betweencomplex
knowledgeand messynarrativemodes. In semioticterms,the sym-
bolic and indexicalrelationships of signification
are reinforcedby
an iconicone. And althoughthereare verygood reasonsto believe
thatcomplexity and messinessare valuableas well as contiguous,
the iconicity in questionis not one of them.Indeed, such a coin-
cidencebetweencontentand form,suchformalcongruence, partakes
of a profoundly mimeticimpulsethatmakesus tend to thinkthat
thereis a virtuein such iconicityin itself.I would like to keep
distrusttowardsuchiconismalongwithan interest in theconnection
itself.
This paper, then, presentsan examinationof epistemological
adventuresin the threeareas in which I have been particularly
interestedsince I startedto workon narrative:anthropology, art
history,and feminist theory.This allowsme,incidentally, to present
the threebestbooksin thesethreeareas I have read in years.Each
of themnotonlydiscussparadigmsof knowledge, butalso constitute
these. Hence, if I eventuallyhave some criticalremarksto make,
the readeris askedto bear in mindthatthoseare quibbles,perhaps
inevitablereservations, whichstand out the more emphatically as
these studies are such exemplary texts which only my unease with
the termpreventsfromcalling masterpieces.To speak withLorraine

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
NARRATIVE AS EPISTEMOLOGY 299

Code's preferredmodel: I willengage these paradigms of knowledge

according to the mode of friendship,not the adversarial mode.
I will discuss the potential of narrativeas epistemologyas well as
the problems a narrativeepistemologymight incur by confronting
views of narrative in these three studies, eventuallyin relation to
time and "person." These two aspects of narrative,it turns out,
have a tremendous impact on the very possibilityof reliable and
responsible knowledge. The coincidence thatboth Fabian and Dam-
isch find their inquiries to converge in the notion of performance
as an alternativeto dominating,exploitative,and asymmetricalmodes
of knowledge requires an examination of the implicationsof that
concept in relation to the narrativeaspects just mentioned. But as
I said before, coincidence itself,with its leaning toward mimesis,
will benefitfrom my continuous doubt.

II. Narrative and Epistemology

Let me firstexplore some incidentaland less incidentalconnections
between narrativeand epistemologyas these studies display them.
Code's firstconcern is to break away from the dichotomybetween
objectivismand relativism,and given my own inclinationto deplore
dichotomies, I was already interested right there. As it happens,
her view mediates between the two opposites by virtueof narrativity.
Here is her definitionof relativism:"Broadly speaking, epistemo-
logical relativistshold that knowledge, truth,or even 'reality'can
be understood only in relation to particularsets of cultural or social
circumstances,to a theoretical framework,a specifiable range of
perspectives,a conceptual scheme, or a form of life" (W 2). While
not endorsing a stark construal of relativism,nor the equation of
epistemological with conceptual relativism,she mentions as major
advantages of a moderate epistemologicalrelativismthe fact that it
"is one of the more obvious means of avoidingreductiveexplanations,
in termsof drasticallysimplifiedparadigms of knowledge,monolithic
explanatorymodes, or privileged,decontextualizedpositions"as well
as the "stringentaccountabilityrequirements"it entails (W 3). And
these remarksnicely sum up the ambitionsof both Fabian and, less
explicitly,Damisch. Both go out of their way to avoid decontex-
tualized reporting,and they do so by experimentingwith narrative
structure.One of the aspects of the theoreticalframeworkand the
conceptual schemes, as well as, in Fabian's case, the "ways of life"
of Code's definition,is later explicated as narrative. Code even
makes narrativethe core of her "epistemicresponsibility."'9
She argues

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
300 NEW LITERARY HISTORY

that the moderate relativism she advocates entails an increased

relevance of narrative: "once epistemologistsrecognize the locat-
edness of all cognitiveactivityin the projects and constructionsof
specificallypositioned subjects,then the relevance of narrativewill
be apparent as an epistemologicalresource" (W 170), and she adds
thatthe model of the Cartesian knoweras neutraland not positioned
has worked to obscure that significanceof narrative.
The importance of narrativebecause of its capacity to map po-
sitioned subjects in relation to knowledge does not entail a facile
rejection of all standards of objectivity.On the contrary,as Code
rightlyargues, while on the one hand, "often, objectivityrequires
taking subjectivityinto account" (W 31), on the other denying that
there are objective social realities "would obliteratethe purpose of
feministpoliticalprojects" (W 45). Rather,the subjectivitiesinvolved
in the interactionsthat especially humanists and social scientists
studyare objectifiablepreciselybecause theycan be related to, made
relative to, and positioned within narrativeconceived as a mobile,
dynamic, conceptual scheme.
As I mentioned before, Code's competitorfor paradigm statusto
supersede physics is "knowing other people," and although she
doesn't name any academic discipline, it seems obvious that an-
thropologyat itsbestcould be the privilegeddiscipline,withnarrative
as its central mode. That the history of anthropology has not
particularlyyielded such statusstemsfromthe bond between knowl-
edge in the objectivistmode and domination as a political practice.
But revised in this direction-and such a revisionis well under way,
with Fabian as one of its leaders- Fabian's study could then be a
paradigm withinthe paradigm. Knowing other people-which, for
Code, is best seen as based on the model of friendship-has features
thatclearlydemonstratewhynarrativeis such an importantresource
for it, and all of these featuresare prominentlyat work in Fabian's
book: such knowledge is not achieved at once, instead it develops;
it is open to interpretationat differentlevels; it admits of degrees;
it changes; subject and object positionsin the process of knowledge
constructionare reversible;it is a never-accomplishedconstantproc-
ess; "the 'more-or-lessness'of this knowledge constantlyaffirmsthe
need to reserve and revisejudgement" (W 37-38). This last feature
pointsat the need forself-reflection as part of the epistemicendeavor
itself.
The relevance of narrativeas a resource is not limited to its use
in documents and reports; the process of knowledge construction,
whichboth Fabian and Damisch like to call performance,is narrative
in nature on all scores. The events that constitute the process

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
NARRATIVE AS EPISTEMOLOGY 301

producing knowledge do not exist outside the narrative accounts

of them,which constructthe knowledge by representingthe events.
Moreover, the knowledge-claimantsposition themselves within a
range of what Code calls "discursive possibilitieswhich she may
accept, criticize,or challenge" (W 122), thus constructingyetanother
performativecontext which does not admit reduction to simples
and separation of discovery,justification,and report.
Damisch paradoxically demonstrates the pervasive relevance of
narrative in his resistance to it when he writes: "That a painting
cannot be narrated is-as you noticed in the beginning-a kind of
scandal in a culture so massivelyinformedby philologyas ours" (L
239). Later on I will revertto the odd bracketedclause in the second
person, but fornow I wish to remarkthatthisstatementis noticeable
by its inherent contradiction:The Originof Perspective is preciselya
narrative of paintings,but of paintings as actions, taking the pro-
gressiveformliterally;"scandal" impliesa storyand, indeed, Damisch
proceeds to devote the rest of his book to the narrativeof the three
paintings he has selected. A narrative more narrative than those
constructedby his fellow historianshe so generously despises, for
it tells the story of the paintings' performance,including various
characters,events,focalizers,and even narrators.His "epistemology
of the group" precisely turns three isolated and perhaps static
paintings into a set of characters among whom events--essentially
relative transformations--takeplace.1'
Code's central critique of the traditional Cartesian subject of
knowledge challenges the individualisminherent in that tradition.
She sharplydenounces the blatanttensionbetween the autonomous,
pure, and unique subject of objective knowledge and the reduction
of people who are "objects" of study to "cases" or "types" (W 21).
Distinguishingautonomyfromindividualism(W 78), she emphasizes
the impossibilityeven to conceive of subjects as individuals inde-
pendent from the senses, the social structures,and other people.
In an argument strikinglyconvergent with that put forward by
linguist Emile Benveniste, she convincinglysuggests that "persons
essentiallyare second persons" (W 82), meaning thatthe dependency
on caretakers and other people makes personhood in isolation
impossible. Language alone, the very language knowledge is so
heavily contingent upon, proves it. Similarly,Benveniste claimed
that the first-personpronoun that produces linguistic subjectivity
can only be semanticallyfilled by a second person acknowledging
and eventually reversing it. And that is the very reason why for
him the pronoun, not the concretenoun, is the essence of language;
deixis, not reference.It is thisdependency on othersthatconstitutes

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
302 NEW LITERARY HISTORY

the scandal, the stumblingblock, of orthodox epistemology; and

hence it is the traces of that grafted status of the knowing subject
that must be erased (W 172). Thus formulated,the problems and
tensions within this epistemology resemble that of "third-person"
narrativein the realisttradition,where subjectivetracesof narratorial
interventionmust often be erased, but must at any rate not be
explicitlyresponding to an implied second person thanks to whose
curiosity,antagonism,or interestthe narratorial"I" can constitute
itself. Solicitationby the second person cruciallydefines firstper-
sonhood; and thereforethe lattermust hide behind impossiblethird
personhood; just as in visual representation,the allusion to per-
spective rather than the full embodiment of it-though an incon-
spicuous allusion-works to both stage and hide the subject."
Fabian and Damisch are quite outspoken in their "second per-
sonhood" and thereby constitutethemselves as ironic, self-aware,
perhaps postmodern narrators. They do that in several manners,
of which more shortly,one of which is to struggle with the very
textualitythey need in order to performtheir knowledge. I have
already quoted Damisch's resistanceto narrative,and Fabian's chap-
ter title"Interlude: The Missing Text" points to a similarproblem.
In it he discusses the differentconceptions and genres of textuality
currentlydebated in anthropology,such as the equation of culture
and text,'2 or the experimental practice of literarygenres in eth-
nography,'3or the literaryanalysisof ethnographictexts.14But these
are conceptions of textualitythat do not affecthis work, his per-
formance, in writing this book. Then he begins to explore the
predicamentof the textson which ethnographersbase theirwriting:
fieldwork notes, documents,recordings,protocols.The tremendous
problem of makingtextto which the chapter's titlenegativelyalludes
ends up beingthe textwe are reading, whichenables Fabian to come
up with the followingirony: "Never before did I have the chance
to witnessand document text production in such detail. But there
is no hope ever to come up with a definitivetext of the play" (P
91). The irony bites itself in the tail when we realize that it is
precisely the story of that irony that we are reading. For Fabian
does narrate not so much the production of the text as his docu-
mentation of that production. Narrative as a mode entails that
inevitablymetanarrativeposition: Fabian cannot perform (his role
in) the collectiveconstructionof knowledgeby a number of different
subjects/characters withoutbeing the narrator-focalizerof the story
of that construction.As a consequence, in spite of the above quo-
tation, what he comes up with is neither the text of the play nor
the text of the text production but, first,a wonderfullyclever

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
NARRATIVE AS EPISTEMOLOGY 303

structuralrepresentationof his focalization of the production of

the former through the latter: the transitionsfrom discussion to
plot design to play-makingon the one hand, and the gradual reversal
of the respectiveamounts of talking and of acting on the other.'5
To understand how this paradox is bound up with narrativeon
more than an anecdotal level, it can be compared to Damisch's
analysis of linear perspectivein terms reminiscentof narratological
typologies of narrative situations. Formulatingthe hypothesisthat
perspective provided painters with a network of indexical signs
equivalent to the systemof enunciationin language, he demonstrates
various possibilitiesof relating to the "law" of perspective,each of
them equally narrative. Either one obeys or ignores the law, in
whichcases two narrativesituationsare unambiguouslyrepresented.
Or one only puts in a sign or two of it, not necessarilycoherent
among themselves,just enough to make the "law" work: to make
it appear to be assumed, endorsed by the viewer. This is how
perspective,even withinthe practice of painting,is a discourse: it
can be intertextuallysignifiedwithoutbeing obeyed and yet it will
be read. This would be as close as one getsto "third-person"narrative
with an invisible narrator. Or a painting can refer to the model,
but only to deny it. Damisch demonstratesthiswithRaphael's Extasis
ofSaintCecilia,where perspectiveis heavilysignifiedyetnot obeyed.'6
Such a denial can work like a self-ironicalstatement.Damisch rightly
adds that,rather than underminingor invalidatingit, such a denial
reaffirmsthe system.
This latter situationcan be compared to Fabian's predicamentof
irony upon irony, when his denial of his narrativecompetence in
fact affirmsit, and my guess is that he knows it. The strugglewith
text-makingis a struggle for the abilityto answer Damisch's very
pointed question, "S'il y a histoire,de quoi est-elle l'histoire?" (L
12; If there is history,of what is it the history?).And this question
is, I like to think,the meeting-pointof narrativeand epistemology.
But this is so precisely because that question does not bear a
simple answer. For Fabian's predicament is precisely that the pro-
duction of the knowledge he wants to narrate is a performancein
which he is an actor, and as I will argue later, in some way he is
its hero. His is, as I said before, a "first-person"narrative,auto-
biographical from the beginning. He needs to act in the text which
he thereforecannot writeup, for performanceprecludes narrative
in the "third person." Of whatis he writingthe history?Of himself
writingthe historyof himself writingthe historyof. . . . Instead
of providing a simple answer to his own question, Damisch's whole
book develops the complex answer which, in the case of the origin

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
304 NEW LITERARY HISTORY

doublesup thesubjectof inquiry.

of perspective, He writesmuch
later in the book a sentence that displays the difficultyin its very
structure: "But there are various ways of conducting a narrative
. . . which does not necessarilyimply the a priori constructionof
a scene,and even less the production-even if strictlyfor the sake
of demonstration--ofan apparatus (dispositif) where representation,
in the modern sense of the word, would be asked to reflectitself
in its operation, and simultaneouslyin its constitutivereference to
the position of the subject" (L 364). If the discourse here seems to
become hopelessly entangled in its subordinate clauses and double
negatives, it is, I think, because Damisch is describing as well as
demonstratinghere how difficultit is to be entangled in the "first-
person" narrator's position of a performance that stages that nar-
rator. The days of Brecht and the epic theater are long gone, and
so is Freud's mysticwritingpad; and whatremainsis the impossibility
of answering the question "Of what is it the history?"upon the
scene of writing."7 The acceptance and handlingof thatcontradictory
entanglementmay well be, at the same time, the crucial relevance
of narrativefor epistemology.

III. Facing Domination

Earlier on in this paper I quoted a statementfrom Code which

suggested that cleanliness,or at least neatness, had a lot to do with
the preference for physicsas the paradigm of knowledge,together
with the resistance to narrative,subjectivityand, as some episte-
mological texts suggest, women as subjects of knowledge. Indeed,
Wilhelm Von Humboldt's judgment that "their [women's] nature
also contains a lack or a failingof analyticcapacity which draws a
strictline of demarcation between ego and world; therefore,they
will not come as close to the ultimateinvestigationof truthas man"'18
may have been replaced with more sophisticatedversions of the
same, and I wouldn't wish to suggest that women have always and
everywherebeen excluded from knowledge. But the particular in-
terestof this text remains in the reason alleged for that exclusion,
and to whichthe word "messy"in Code's statementresponds. Indeed,
feministshave amply demonstrated the vested interestof a "male-
stream" view in the securing of boundaries, of countriesas well as
bodies and intellectualterritories.And the projectionof the violation
of those cherished boundaries onto those subjects who, according
to a biological iconism, are subject to it speaks of the conceptual
and emotional confusion underlyinggynophobia. The confusionof

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
NARRATIVE AS EPISTEMOLOGY 305

subject and object, only too well known,which underlies this phobia
happens to be a powerfulideologeme, or even an ideological code,
serving many purposes, and we will encounter it once more in the
present inquiry."9This is one reason why a subject-orientednar-
ratologycan be helpful.20
The investmentin boundaries-here you have a subjective,emo-
tional motivationforobjectivity--Codesuggested,enhances the need
for observational simples as the basic unit of knowledge. General
epistemologythus partakes of another specificideological code, that
of the accumulative principle in the name of which many scholars
claim that objects consist of the sum of their parts. This principle
hampered the developmentof semanticsuntilthe advent of discourse
analysis,forexample. As a thirdparticipantin thisideological cluster,
we mightcount dualism, not onlythe mostbasic structureof Western
thought but also the all-but-exclusivemode of academic argumen-
tation. Here yet again the structurelives off the artificialand often
unwarranted isolation of well-delimited(boundaries!) claims and
arguments.The mode produces less than maximallygood reasoning,
as Janice Moulton, who came up with the concept of adversarial
mode of argumentation,argues, since it excludes both complications
of the issue when taken in context and plural approaches to it.21
Given the need for sharp opposition and delimitationthatthe mode
demonstrates,it is structurallycomplicitouswith objectivism,which
depends on equally strongdistinctions.In other words, the subject-
object distinctionof objectivismis structurallysimilar to, and con-
tiguous with,the self-otherdistinctionof the adversarial mode. And
since it uses the model of war for the peaceful activityof intellectual
work, what it also betrays is the intricate relationship between
knowledge, aggression, and domination.
demonstrated,by EvelynFox Keller
Indeed, it has been sufficiently
and others,that the sharp divisionbetween subject and object which
encourages adversarial attitudes is predicated upon the implicit
notion that the goal of knowledge is "to produce the ability to
control, manipulate, and predict the behavior of its objects" (W
139).22 The obvious question, then, becomes, What are the stakes,
and why are these so high as to entice well-meaning,serious, and
self-confidentscholars to cling to a model so contaminated by
objectionable impulses? Keller looks at psychoanalytictheoryfor an
answer, and she makes her case with much force. But withinthe
present inquiry such answers tend to beg the question. If strong
boundaries provide emotional comfort,and if that is so especially
forsubjectswho need thatcomfortmost,the cultural,representational
forms that scheme takes remain to be interpreted.

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
306 NEW LITERARY HISTORY

Damisch provides an element for an answer in his analysis of

perspective as just such a device for demarcation. He has a keen
sense of the issue when he writes: "In order for the thingsin this
world to become objectsforperception,the subjectmusttake distance
from itself. . . . But that movement, even in its slight theatricality,
remains subjected to the law which is the law of representation:the
distance the subject takes in relation to the object . . . allows him
to escape to the immediatelylived experience; but he can only
discoverthathe is implicated,irremediablyso, in the spectaclewhich
takes its truth from that very implication" (L 345). This implicat-
edness which is the very essence of the systemof perspective as
well as its motivationhelps Damisch to understand the "difference
within"perspectiveas illusion, bound up with realism but not with
reality,a provider of the illusion, precisely,of original subjectivity.
Ironically,the subject who needs to see its origin mirrored in the
systemof perspective,"thatsubject which is considered 'dominating'
since it appears to be established in a position of domination is
tenuously established [ne tient qu'a un fill" (L 354).
Domination,then,is not a politicalbackgroundof representational
realism but its product. Yet at the same time that product is illu-
sionary, imaginary. Much earlier, Damisch had quoted Merleau-
Ponty,who equated such a mode of vision withdomination,illusion
("the inventionof a dominated world" [L 46]), and adulthood, and
then he had continued: "A vision in the firstperson, coherent,
mastered, and which would imply as its condition the position of
a subject who can eventuallyclaim it as his, as his property,as his
representation"(L 46). In addition to the ideological problematic
this statementimplies, there is one confusingepistemologicaldetail
here, which is related to one we saw earlier without stopping to
consider it. The juxtaposition of "his property"and "his represen-
tation" points again to a confusion of subject and object, if not to
a mechanism of projection. If the visual fieldencourages subjects-
adults, according to Merleau-Ponty--totake hold of the objects in
that field, to deploy the gaze, if we think of Norman Bryson's
distinctionbetween gaze and glance,23they mighttend to consider
it their property.This would place an urgencyon the debate about
pornography,for example. But what about the ambiguous phrase
"his representation"(sa representation):Does it mean the represen-
tation of which he is the object or the representationhe performs,
through the illusionarymirroringprovided by the deceptive optical
structureof perspective?If so, the fantasyof adult vision as a first-
person narrativeforegroundsthe ambiguityof preciselythatnotion.
For we have seen that perspective, with its smooth if illusionary

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
NARRATIVE AS EPISTEMOLOGY 307

effect of the real, works precisely because it both inscribes and

effaces the subject of vision.
But the fundamental confusion that underlies the equation of
speech and the look in a speech-act-orientedtheory of vision is
preciselythat same illusionaryorigin Damisch's entire book works
to explain, yet reaffirmsin this theoreticalmoment. For the subject
of vision is notthe subjectof paintingbut its addressee. Firstperson,
second person? Are these, in effect,the same person, and what
would the consequences of such a conflationbe for epistemology
in general, for Damisch's writingstyle,and for Fabian's project of
a critical,communicative,dialogic, performativeanthropology?The
question is relevantin the lightof the obvious struggleboth writers
are engaged in, which is a struggle explicitly to do away with
domination by doing justice to the second person.

IV. Second Person?

The concept of second personhood has, then, a triple allegiance.
First, as presented by Code, it indicates the derivative status of
personhood; the fundamental impossibilityto be, both psychologi-
cally and socially,a person withoutthe traces of the person's grafted
being. Second, as presented by Benvenisteand subsequent theorists
in his vein,it indicatesthe reversiblerelationshipof complementarity
between first and second-person pronouns whose use produces
subjectivityand constitutesthe essence of language precisely,Ben-
veniste says, because the pronouns do not refer.24Note that both
these allegiances are defined negatively,underminingthe humanist
individual who ruled over objective knowledge,the knowledge that
effectivelyhad an "object." Third, then, it indicates the partner of
the ethnographer and the historian, those persons, subjects, or
discourses formerlyreferredto as the "object" but now engaged in
the dialogue of the performance. To these second persons, the
scholars have a strong allegiance that is both epistemological and
political. But to avoid the traps of ethics in the overextended use
of the political, I will just use the former term. Narrative, as a
structuralformand as a discursiveposture, presentsa unique place
to study the intertwinements of these three allegiances. In such an
analysis lies perhaps the most valuable epistemologicalcontribution
of narratology.
Narratology is the theory of narrative,and it provides tools for
analyzing narrativetexts. A workingdefinitionof narrativemay be
in order here, to avoid both overextendingand needlesslyrestricting

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
308 NEW LITERARY HISTORY

the concept. A narrativeis an account, in any semiotic system,of

a subjectivized and often entirelyor partly fictionalizedseries of
events. It involves a narrator--whetherexplicitlyor implicitlyself-
referential,always a "firstperson"-a focalizer-the implied subject
who "colors" the story-and a number of actors or agents of the
events.Narrativethusconceived is not confinedto literaryor, indeed,
verbal narrative. It is a mode of semiotic behavior rather than a
finiteset of objects. One aspect of that semioticbehavior is the one
under scrutiny here: the use of first, second, or third-person
discourse.
This psychosocial,linguistic,and epistemologicalsecond person-
hood affectsboth parties, the "firstperson," subject of inquiryand
writing,as much as the second person, the interlocutorsand fellow
inquirers in Fabian's case, and the historical"other" discourses in
Damisch's case. First and second-person positions are by definition
reversible,and one way to measure the success of this epistemic
styleis preciselyto examine the actual reversibility.From now on,
I will treat these texts as literarynarratives,worthyin themselves
of detailed analysis. And as happens in such cases, the analyst can
only point at a few exemplary features and details, not be com-
prehensive at all. In Fabian's case, the structuralproperty of the
text I will focus on is the narrativestructureof embedding and of
the representationof "characters."For Damisch, I will look at the
microstylisticfeature of the use of grammatical"person," especially
in the second part of the book. Throughout this analysisI will keep
connecting narrativestructureand epistemic meaning.
Fabian's beginning has been quoted already. It sets him up as a
first-person narrator-character, engaged in reportingeventsthatcan
be summarized as "his discovery."The story of the discovery is
gripping: at the punctual moment of the evening in 1986, the
narrator realized that the interpretiveevents around the proverb
"power is eaten whole" constitutewhat he names "a new ethnog-
raphy." Thus the anecdote of the discovery attributedto the "I"
appears as a frame narrative,embedding a second narrativewhich
elaborates the circumstancesof the discovery,the narrativeof the
anecdote of being told the saying.
The structureof embedding is importanthere. In the first-level
narrative,the narrator is the firstperson, and appropriately,he is
on his own. In the second-level narrative,the embedded one-but
the structurewill not remain so neat-the narrator appears as a
second person, being told, by his Shaba interlocutors,somethingin
plain words that he does not understand,but upon which he needs
to act culturally "correctly."Like in Gide's Faux Monnayeurs(The

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
NARRATIVE AS EPISTEMOLOGY 309

the modernistquest for meaning begins here.

Counterfeiters),
The second-levelnarrativeof the quest for meaning of the saying
overflowsinto the firstlevel when its provisional denouement rep-
resents the shared ignorance of Shabans and expatriate ethnogra-
pher, resultingin the brainstormingsession that is the startingpoint
of the experiment.Given the delicacyof the exchange, the inequality
of knowledge-even if they cannot interpretthe saying,the inter-
locutors "know" it better than Fabian-the problem that it is the
first-personnarrator who is telling both tales, and the intricate
narrativestructureof the overall text,it is relevantto ask in which
direction that overflowends up streaming.
But there is yet another level. Woven through this narrationare
reflectionson ethnography.At firstsight these are discursive in-
terludes,argumentativein mode, articulatingan argumentas distinct
from a narrative which represents a story. Yet they are in turn
narrativizedas Fabian's personal quest for the best method during
the past ten years. His cherished dialogic, communicativemethod
had given him pause already, he tells us, firstbecause of its false
ethical suggestion of equality, hence its illusive righteousness,and
second because, epistemologicallyspeaking, it does not enable one
to account for the production of knowledge. Since the dialogic model
assumes that knowledge is shared, conveyed by those who have it-
the membersof the culture being studied-to those who desire it-
the ethnographer-it begs the question of how the knowledgecomes
about.25
Thus we have three levels so far: the punctual, first-personstory
of the discovery; the storyof the evening of the proverb dinner,
continued during, say, a few weeks of search for meaning, ending
in the group of actors who stage the saying, told in a first-second-
person dialogue with reversible positions; and the story,again in
the firstperson but withan implied second person--Fabian himself,
at an earlier moment,as well as his fellowanthropologists-running
through ten years but interspersedwith many "achronies."26
Fabian is an engaging narrator,and his text is so explicit in its
epistemic position, as well as overt in acting that position out, that
it takes a second look at the overall structureof the text to realize
a potential problem. A problem that,it is only fair to say, he could
hardly have avoided, and which by no means undermines his tre-
mendous accomplishment.Yet the problem is major: by virtue of
the very narrative form, the second person cannot but be subor-
dinated to an extremelyself-centeredfirstperson.
Indeed the text as a whole mirrorsthe structureI just outlined
for the firstfew pages. Chapters 1 through 5 are primarilya first-

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
310 NEW LITERARY HISTORY

person narration,embedding the multiple narrativescharacterized

as second-level-embedded,second-person narrative,and a laterally
connected, partlyalso second-level,first-personargumentativenar-
rative with a stronglyimplied second person identical to the first
person. Fabian deploys manystrategies,some of whichare extremely
effective,to empower the embedded second persons. Thus, for
example, in the thirdchapter he provides a shorthistoryof theater
in Shaba, and of the Mufwankolo group in particular,in which he
is careful to furnish,in footnotes,individualized life-historiesfor
all characters mentioned. While this would be a troublesome kind
of individualistichistoriographyin a Westerncontext,here it serves
the emancipatory purpose of individualizingpeople so far mostly
seen as ahistorical"folk."
Chapters 6 through 13 constitutethe ethnographyproper. Here,
the second persons-the group of Shaba actors-are the principal
speakers. Fabian is meticulous in doing his utmostto enable these
speakers. This part has, again, three forms. First,the text is tran-
scribed in Shaba Swahili. Second, the English translationfollows,
symbolicallyin the second place. Third, both versionsare provided
with helpful footnotes,clearly meant to be subservientto the en-
terprise of opening up the main, second-person text. This text is
"second person" in two senses: it is the text produced by the second
persons, Fabian's interlocutors,and it is dialogical in kind itself,
since it transcribesthe dialogues that took place in the construction
of the play. In this part, the second persons remain in firstposition;
in spite of the factthatthe bulk of the transcribedrecordingsmight
seem in need of an explanatory,interpretive,academic commentary,
relegating this commentaryto footnotes is a rhetorical means of
effectivelypreservingthe primaryposition for the Shabeans.
The concluding chapter is, again, writtenin the firstperson. This
text has a metapositionin relation to the second part as well as the
first,while it is also a continuationof the argumentativeinterludes
in the beginning. The second person of this third part is clearly
the "Western"anthropologist.Thus a formulationlike the following
strikesme as out of tune with the careful narrative-epistemological
strategiesof the firstand second parts: "First,it is wrong to assume
that the Zairean 'folk' . . . live only in the present and, as folk are
said to do, only worry about forms of power and oppression as
they exist now" (P 286). Whereas this passage pointedly opposes
mistaken and yet tenacious prejudices, and thereforeis obviously
veryuseful,it cannot help but statethe "truth"about Shaba Zaireans
who are thus relegated to third personhood. And this happens in
the terms,albeit bracketed,of the Westernoppressive heritage.And

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
NARRATIVE AS EPISTEMOLOGY 311

I don't mean just the use of the term "folk" but, more insidiously,
the very fact that the passage responds to a judgment couched in
the categories of Western philosophy: time, present, history.
From the vantage point of this final part, then, the text can be
seen in the lightof Lejeune's analysis of autobiographyin the third
person: "Dialogue. Now the aim is not to constructbut to destroy a
point of view towardoneself. The dialogue is presentedas a response
to a discourse already expressed but which must be reconstituted
for purposes of refutation.This earlier discourse will be reenacted
so that it can be answered. In the frameworkof an autobiographical
text presented as such, a fictivetrial is thereforereproduced; pros-
ecution and defense are set up and allowed to speak. Of course,
the discussion soon favorsthe autobiographer,who graduallyallows
his true image to emerge victorious.""27
And indeed as a consequence of the tripartitestructureof this
book, one may want to look again at the ways the second person
has been staged in thiscomplex narrative.Embedded in a masterful
and masterlyfirst-personnarrative,the Shaba actors end up serving
the interest of substantiatingFabian's discovery. This discovery,
moreover,concerned less the knowledge produced about the Shaba
insightsinto power, than Fabian's insightsinto his discipline. Self-
reflection,however indispensable, sometimes courts self-centered-
ness. And whether this danger becomes a serious threat depends
on the interplaybetween firstand second personhood. In this case,
the narrative structureof the text, both globally and in detail as
analyzed for the beginning, suggests that the second person has
been subsumed under the first,therebylosing if not its alterityat
least its power to put that alterityfirst.And as we will see shortly,
this subsumptionis reinforcedbecause it also takes another form-
thatof mimeticism,already alluded to in the beginningof thispaper
a propos of Code's messy narrativesfitfor complex ideas. But let
me turn to Damisch's narrativefirst.
The dubious statusof the second person becomes far more blatant
in Damisch's case. Whereas he theorizessecond personhood through-
out his book as part of the problem (the "object") he is analyzing,
epistemologicallyhe limitsit to a rhetoricalstrategywhichhe imitates
from the ancient treatiseshe studies. This book is explicitlydivided
into three sections. While the entire book carries along, in parallel
and intertwinement,the epistemological debate addressed to the
writer's fellow historians with the analysis of the historyof per-
spective, this discussion receives primary focus in the firstpart,
following up on the initiallystated "impatience." This firstpart
elaborates Damisch's challengingview that perspectiveis a discursive

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
312 NEW LITERARY HISTORY

apparatus of enunciation (L 38) based not on the fit but on the

mismatchbetween geometricaland symbolicpoint of origin (L 56).
It is that mismatchthat produces visual subjectivity.
The second part engages the ancient treatisesand their writers
as the second person. These second persons become firstpersons
in a real sense in the long, often full, quotations-equivalent to
Fabian's second part with the full Shabean texts. These fragments
are quite thoroughlyinterpretedand addressed, from the point of
view of the modern scholar,who therebyacts as the second person
responding to firstpersons.
It is the thirdpart which is both the most importantand presents
the most problematicversion of second personhood. It contains the
actual analyses of the three perspectivalpaintingsin relationto one
another. Here the author elaborates the epistemologyof the group
a la L6vi-Strauss,but then historicizedthroughfurthercomparisons,
most notably with Van Eyck's ArnolfiniWeddingand Velaquez's Las
Meninas, up to Picasso's response to the latter.
Much to this reader's surprise,this third part opens withthe use
of the second grammatical person, which we already saw in an
earlier quote. The tone changes, the narrator seems to raise his
sleeves to go reallyto work,and here is how he justifiesthe rhetorical
shift: "And now, this painting. This painting that you know better
than anyone: which forces me, at thisjuncture, to call upon your
testimonyand to shift-according to a device frequentlyused in
the old treatises--fromI to you,and from one discursiveregime to
another,to an explicitlydialogic one" (L 157). The "you" comes up
at the moment that the narratorbegins to tell the storyof his own
engagement with this painting. The rest of the paragraph further
explains the point of this device. Not only does the narrator wish
to pay homage by imitatingthem to the discursive habits of the
ancient writers,his previous second persons. Also, he intends the
pronominal formto signal that "one cannotjust put such a painting
at one's disposal as one wishes, and like a random object or doc-
ument" (L 157). This paragraph is followed by a page and a half
of description,in the third person, of the painting presumablyas
I/you see it. As he will warn us later, in the passage I quoted at
the beginning,description is always already narrative,and in fact
this descriptionof a stillpaintingwithoutany figuresor movement
is a masterpiece of narrativizeddescription.What it narrativizesis
preciselyperspective:"an urban site fixedwithina perspectivewhich
unfoldsbefore the eye the symmetricalfan of its vanishinglines" (L
157; my emphasis).
If we take the use of the second person at the letter,the "you"

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
NARRATIVE AS EPISTEMOLOGY 313

is called upon as a witnesswho is therebyauthorized as an expert:

"this painting you know better than anyone." This expert is then
the focalizerof the descriptionto follow,so that the descriptionnot
only narrativizesperspective and the eye before which it unfolds,
but firstand foremost-on a higher narrative level-the expert
witness focalizing it. And this second-person expert is Damisch,
dissociated from the first-personnarrator to gain more authority.
The use of the second person varies greatly, to the point of
inconsistency.Sometimes the status of "you" as the expert directing
the writingsubject "I" is made more explicit,as in "If you insisted
that we exposed this thesis in some detail, it is because it has been
so badly received" (L 180). At other times the identityof "I" and
"you" is emphasized on an emotional basis: "the only question which
matters to us, after all, to you as to myself" (L 182). But if the
split between firstand second person can be thought to signifythe
differentfunctionsof narratorand focalizer/expert witness,at other
moments these two functionsare conflated so as to evacuate the
point of the linguisticgame: "There is still a problem you have
already mentionedonce or twice"(L 249), where "you" incongruously
is the writingsubject/narrator.In the end, it seems "you" and "I"
overlap completely:theyhave not only the same identity-the same
person in the psychosocial sense-but also the same function,the
same linguisticperson. What, then, is the point of the game, one
may well ask?
The connotativeeffectsof this rhetoricalstrategyare varied, and
do not always overlap with the narrator'sstated intention.To assess
these it is imperative to take into account the other part of the
device, which is the use of a third person. This third person is not
the painting/"object" but the contradicteurs.
By this term the narrator
sets up as diegetic charactersin the wake of the rhetoricaltradition
of which he is writingboth the analysis and a pastiche, the implied
opponents who were present from the beginning,namely his fellow
historians.These charactersappear ratherlate in the day, as Damisch
franklyadmits (L 385). But what interestsme in that appearance
is the rationale they are in charge of offeringfor the pronominal
game as a whole. For Damisch introducesan explicit"third person"
withan epistemologicalaim. No more than Fabian, but foraltogether
differentreasons, is he content with the mere dialogical form of
writing:"As if dialogue did not sufficeto give the debate its true
dimension, and one had to appeal to a third person to put it in
perspective" (L 385).
In an explanation presented on the mode of fictionality ("as if")
and in a stronglyvisual vocabulary,the narratorjustifieshis use of

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
314 NEW LITERARYHISTORY

pronouns in a combination of a truth claim ("true dimension"), a

move of distancing(now the thirdperson is called upon as a witness),
and a mimeticact (perspective on perspective).
The effectof the pronominal game stands out most stronglywhen
the three grammaticalcharactersappear on stage together,as hap-
pens, for example, on page 386: "But one/I [on] can respond
differently to the objectionattributedto the contradicteur
(an objection
you are far from taking lightly."The structureis clearly mobilized
for a defensive purpose. The depersonalized firstperson (on) is
going to refute an objection he came up with in the firstplace but
which he attributesto his third person; the second person, the
expert/authority, is said to take the objection extremelyseriouslyso
that the third person has to be satisfied.But since the firstperson
comes up withthe objection,we mustconclude thatthe thirdperson
too is identical to him.
Damisch needs three persons, he claims,because the debate needs
to be put into perspective.Perspective,on the otherhand, is precisely
characterizedby the deceptive illusions of true, neutral,objective-
in other words, "third-person" - representationof the world. Yet it
works so effectivelybecause at the same time it provides the viewer
witha positionas the firstperson who "owns" thatworld. As Damisch
brilliantlypoints out, perspectivesets up the elision of the subject-
tenuously inscribed already-in the viewpointwhich is seen as the
origin of subjectivity.And that elision is signifiedas apostrophe(L
402), enforcinga second person subsumed withinthe firstperson
who otherwisewould remain unsustained.
Thus the rhetoricof thisthirdpart resembles,mimes,its cognitive
content. This is never spoken out but alluded to, tongue-in-cheek,
if only by thejuxtaposition of passages about the one and the other.
But another congruence is more explicitlystated. Toward the end
of the book Damisch seems deeply gratifiedwhen he is able to
suggest thatthe three points involved in perspective-the viewpoint,
the vanishing point, and the distance point--correspond to three
locations: here, there, and yonder. A bit later he then writesthat
perspectiveas a paradigm, as a model that projects,does more than
pose the other in frontof the subject as always already there before
him; it also introduces a "third person" (un tiers).What emanates
is a triangularvisual regime that corresponds to the Lacanian (law
of the) fatherwho comes to break the untenable dualityof mother
and child, wherein the mother cannot be the other because the
third person is needed. This is, it sometimesappears, also the law
of the excluded middle, the principle of dualism.

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
NARRATIVE AS EPISTEMOLOGY 315

V. The Seduction of Mimeticism

Why is it that this argument is more than persuasive--almost

irresistible?I have already alluded to one troublesome feature in
both these books, also present here and there in Code's: the
occurrence of congruence, of a mimeticism.This happens on many
levels, and firstof all on the level of overall structure. Fabian
proposes performance as a method of ethnography "with" the
people described, and as if by chance the object, occasion, event
to be studied is a performance,in a sense that makes the method
appear more "real" than the narrative structure suggests it is.
Damisch uses a triangular rhetoric which substantiates,and is
substantiatedby, his theory of triangular perspective,and only if
taken seriously-as "not a game"-does the rhetoric alert us to
the potential collapse of third and second person into first-just
like in perspective.The mimeticimpulse, once noticed, is pervasive
in both texts. Fabian writes,for example: "It occurred to me that
the group's work-giving form to everyday experience in the
urban-industrialworld of Shaba and thereby making it possible
to reflectand comment on it-was not in essence differentfrom
my own groping for an ethnography of work and language" (P
42). And these coincidences also happen within the actors' own
lives: "their own progress from childhood to mature age coincides
withthe emergence of popular theateras a childrens'entertainment
and its development to present levels of virtuosityand mass appeal"
(P 43). Formulationsto thiseffectare many: "It is also an interesting
document about 'documentation"'"(P 50n. 24): the Zaireans talk
like Europeans about Zaireans (P 69); within the play, "the idea
of mediation and the riskof corruptionwere expressed dramatically
by locating the most serious threat to the chief's power in the
corruptness of the notables,his intermediaries"(P 282).
Once one is alerted to this tendencyto present analogies, and to
present them as positivein and of themselves,it becomes clear that
coincidences of historiesmay well be an added attractionof coeval-
ness, so stronglyargued for in Timeand theOther.Sometimesit even
seems as though the performance circles around one great epis-
temological goal: to become an allegory of "good" scholarship.The
termsof scholarshipare used to describe the play: "the more direct
threat . . . caused by partiality and distortion when it comes to
interpreting"(P 282), whereas "power must be based on true knowl-
edge and supported by people of integrity"(P 282). Hence not only
is the group's performance an allegory of the ethnographer's ar-

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
316 NEWLITERARY
HISTORY

gument for performance as method, but the very content of the

play allegorizes the scholar whose discovery,afterall, it is called to
illustrate.
Damisch's mimeticismshave been pointed out already. He too
suggests an allegorical identificationwhen he defines painting,his
"second person" par excellence, as something"qui donne a penser"
(which makes you stop and think) (P 289)-just like philosophers.
In remarkslike these, he forgetsthe difference of paintinghis whole
book tries to found, and makes painting be a bit too much like
language. This is, I contend,whyhe is unable to see, in thisotherwise
extraordinarilyclever argument,that the speech act theoryof paint-
ing is ultimatelya language-centered analogy; a product of the
mimeticimpulse.
Since this analogy is extremelycommon in the semiotic analysis
of visual art in the line of Benveniste, especially in the work of
Louis Marin, it seems useful to spell the problem out.28 In its
simplestform,thisanalogy is untenable for two reasons. It conflates
differentmodes of perception withoutexamining the implications
of that conflation--thinkingand seeing; speaking is hardly an act
of perception-and it conflatesdifferentsubject positionsin relation
to acts. Visually representing,not seeing, would be the act parallel
to speaking.29Because of this problem, that confusion ruins Dam-
isch's argument and doubles up his rhetoricalmixture of persons
under his own identity.For the point (pun intended) of perspective
is preciselythat very confusion,but then in the other direction. If
it elides the subject under apostrophe, the second person wins out.
And rightlyso: if enunciation can be a model for perspectival
painting, then the viewer acts, but as addressee. Far from "speak-
ing"-the painting does that-the viewer acts, possiblybut not nec-
essarilyactively,as second person. And thatmightwell be intolerable
for the "you," that fake second but in fact authorized firstperson,
who knows the Urbinate painting better than anyone.

VI. Conclusion
But Fabian also writes' propos of theaterin Shaba that "mimesis
had opened a battle ground" (P 56), and if that is so, then it may
also be one within these texts. A battleground, that is, where a
struggleis foughtbetween two contradictoryimpulses: to construct
knowledge in an engagement with the other, and to subordinate
that other once more. That battlegroundcan host fierce struggles
when the issue is "knowledge of other people" on the model of

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
NARRATIVE AS EPISTEMOLOGY 317

friendship yet narrativizedin a first/third person narrative,albeit

sophisticated and dynamic. For narrative as well as epistemologyis
overdeterminedby its traditionsand histories,one of which is the
central position of the knower/narrator.
The analysis presented here is not meant as a review of the
respectivemeritsand flawsof these studies. If it came to evaluating,
the apparentlygreatersuccess of Fabian over Damisch could simply
be attributedto the differencebetween contemporaryand historical
objects of inquiry, and between linguisticallyaccessible and mute
interlocutors.Damisch ultimatelydoes not have a second person,
one could object to mycriticism.Yes he does, I would argue: himself.
Preciselybecause his narrativegame enables him to deny his own
secondariness in the face of the paintings as well as the treatises,
he can get away with ignoring the paintings' first personhood.
Conversely, Fabian's dialogue, more "real" because he can really
talk with the Zairean actors, is, epistemologicallyspeaking, no less
a sham. From his positionalityas a narrator,he struggleswith his
ignorance, and that positionalityenables him to ask questions in
order to alleviate that ignorance. That is not necessarilythe same
as producing (his) knowledge withthem. To put the cards on the
table with still more explicitness: if you look to blame, I am not
sure I would blame either Damisch or Fabian. As White, Kellner,
and Ankersmithave argued for historywriting,the shape of the
storyyou tell determineswhat knowledge you produce. The result
of the above analysis partlyconverges with this notion, but partly
also complicatesit. For the shape-the dialogue, the performance-
could not overrule the mode: narrative.
In the face of the narrative mode, "friendship"may be a good
model only to the extent that it elaborates and refines what the
antagonisticmode of argumentationsimplifiesand obscures. Taken
too literally,or at face value, to use Damisch's visual vocabulary,it
obscures the dissymmetrythat allows the second person to "be
disappeared" yet again. Damisch's beautiful analogy between the
three points involved in perspective and the three grammatical
persons involved in narrative,and the three locations involved in
spatial organization,could do, by way of caution, with yet another
triangle. I am referringto GayatriSpivak's distinctionbetween self,
self-consolidatingother,and absolute other,translatedbyJohn Bar-
rell as "this,that,and the other.""3This absolute otherseems implied,
feared, and then cast out by Damisch's dramatizationof the "third
person" as a projection of an opposition he is still able to master.
Second personhood, in all three senses distinguishedabove and
integrated as they are in narrative,can easily become self-consoli-

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
318 NEW LITERARY HISTORY

dating ("that" helping the firstperson along). This cautionarynote

leaves unchallenged the need for self-reflection.A self-reflection
whichpartakesof a projectthatis politicalas much as epistemological
requires a sharp analysis,not only of intentionsand methods, but,
more importantlygiven the pragmatic nature of language, of nar-
rative.Narrative,as it turns out, not surprisingly,is telling.

UNIVERSITY OF AMSTERDAM

NOTES

1 Johannes Fabian, Powerand Performance: Ethnographic Explorationsthrough Proverbial

Wisdomand Theaterin Shaba, Zaire (Madison, Wis., 1990), p. 110; hereaftercited in
text as P.
2 Hubert Damisch, L'originede la perspective (Paris, 1987), p. 239; hereaftercited
in text as L. Here and elsewhere, unless stated otherwise,translationsare my own.
3 Lorraine Code, WhatCan SheKnow?Feminist Theory and theConstruction ofKnowledge
(Ithaca, N.Y., 1991), p. 139; hereaftercited in text as W.
4 Johannes Fabian, Timeand theOther:How Anthropology Makes itsObject(New York,
1983).
5 Hubert Damisch, Thdoriedu nuage: Pour une nouvellehistoire de l'art (Paris, 1972).
6 The narrativenature of historiographyhas been the object of analysisfor a long
time now, since Hayden White began to explore the rhetoricof historywriting(see
Hayden White, Metahistory:The HistoricalImaginationin Nineteenth-Century Europe
[Baltimore, 1973]). Recent analyses of interestin this area include Hans Kellner,
Language and HistoricalRepresentation: GettingtheStoryCrooked(Madison, Wis., 1990)
and F. R. Ankersmit,NarrativeLogic: A SemanticAnalysisof theHistorian'sLanguage
(The Hague, 1983). While these studies offer useful insightsinto the problematics
of representationin history,they pursue a goal altogether differentfrom mine.
From my perspective it is problematic that they tend to lack a specificconception
of narrativeas well as an epistemologyagainst which to measure the consequences
of their findings.
7 For an analysis of Descartes's anxieties and the way these informed his episte-
mology, see Annette Baier, "Cartesian Persons," in her Posturesof theMind: Essays
on Mind and Morals (Minneapolis, 1985), pp. 74-92. On the influence of language
on Descartes's thought, see Alasdair MacIntyre, "Epistemological Crises, Dramatic
Narrative,and the Philosophy of Science," The Monist,60 (1977), 453-72. For the
feministimplicationsof this typical mode of thinking,see Susan Bordo, The Flight
toObjectivity:
Essaysin Cartesianismand Culture(Albany,N.Y., 1987) and "The Cartesian
Masculinizationof Thought," Signs, 11 (1986), 439-56.
8 Philippe Lejeune, "Autobiographyin the Third Person," tr. Annetteand Edward
Tomarken, New LiteraryHistory,9 (1977), 27-50. Lejeune writesof Rousseau, that
exemplary first-personwriter'sautobiography in the third person, "He gives us a
lesson in objectivity"(45).
9 The allusion is to Lorraine Code, EpistemicResponsibility (Hanover, N.H., 1987).
10 The "epistemology of the group" clearly shows structuralisttendencies, and
sometimeseven the formulationsrecall Claude Levi-Strauss,especially The Raw and
the Cooked:Introduction to a Scienceof Mythology I, tr. John and Doreen Weightman
(New York, 1969). For an analysisof Levi-Strauss'sconcepts and method used there,

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
NARRATIVE AS EPISTEMOLOGY 319

see Jacques Derrida, "Structure,Sign, and Play in the Discourse of the Human
Sciences," in Writingand Difference, tr. Alan Bass (Chicago, 1978), pp. 278-93.
11 I hesitate to propose this analogy, for reasons I will later expose. I do not
believe it is right to equate the subject of speech with the subject of the look, but
this is as yet another problem of second personhood.
12 See CliffordGeertz, The Interpretation of Cultures(New York, 1973).
13 See George Marcus and Dick Cushman, "Ethnographiesas Texts,"AnnualReview
of Anthropology, 11 (1982), 25-69.
14 See WritingCulture: The Poeticsand Politicsof Ethnography, ed. James Clifford
and George E. Marcus (Berkeley, 1986).
15 See Fabian, Powerand Performance, tables on pp. 93 and 94.
16 Raphael, The Extasisof Saint Cecilia,ca. 1515-16; Bologna, Pinacoteca nazionale.
See L, pp. 38-40.
17 The allusions are to Sigmund Freud's shorttext"A Note Upon a MysticWriting-
Pad" (1924), The StandardEditionof theComplete PsychologicalWorksof SigmundFreud,
ed. and tr. James Strachey (London, 1953-74), XIX, 227-32, and Derrida's com-
mentary"Freud and the Scene of Writing,"in Writingand Difference, pp. 196-32.
These allusions are not just playful; both texts deal with the difficultyof writing
and reading that Fabian is contending with.
18 HumanistwithoutPortfolio:An Anthology of the Writingsof Wilhelmvon Humboldt,
ed. Marianne Cowan (Detroit, 1963), p. 349, quoted by Code, WhatCan She Know?,
p. 10.
19 The termideologeme is borrowed from FredricJameson, The PoliticalUnconscious:
Narrativeas a SociallySymbolic Act(Ithaca, N.Y., 1981). Ernstvan Alphen has theorized
ideology as a code ratherthan a semanticunit (Ernst van Alphen, Bang voorschennis?
Inleidingin de ideologiekritiek
[Utrecht, 1987]).
20 See Mieke Bal, Narratology:Introduction to the Theoryof Narrative,tr. Christine
van Boheemen (Toronto, 1985), fora textbookversionof such a theory,and "Narrative
Subjectivity,"in On Story-Telling: Essays in Narratology, ed. David Jobling (Sonoma,
Calif., 1991), pp. 146-70, for a discussion of the importance of the subjectivity
network.
21 See Janice Moulton, "A Paradigm of Philosophy: The Adversary Method," in
Sandra Harding and Merrill B. Hintikka,DiscoveringReality:FeministPerspectives on
Epistemology, Metaphysics,Methodology,and Philosophy ofScience(The Netherlands,1983),
pp. 149-64.
22 See Evelyn Fox Keller, Reflections on Genderand Science(New Haven, 1985).
23 Norman Bryson,Visionand Painting:TheLogicoftheGaze (London, 1983); further
theorized in Mieke Bal, Reading "Rembrandt": BeyondtheWord-Image Opposition(New
York, 1991), ch. 4.
24 Emile Benveniste,"Subjectivityin Language," tr. Mary Elizabeth Meek, in Critical
TheorySince 1965, ed. Hazard Adams and Leroy Searle (Tallahassee, Fla., 1986), p.
730.
25 This problemis connected to that,addressed by Geertz,in his distinctionbetween
experience-nearand experience-distanceconcepts(CliffordGeertz,"From the Native's
Point of View: On the Nature of Anthropological Understanding" in his Local
Knowledge:FurtherEssaysin Interpretive Anthropology [New York, 1983], pp. 55-70) as
well as to that, discussed by Turner, of the question when and to what extent the
members of the culture are the most adequate informants(Victor Turner, The Forest
of Symbols: Aspectsof NdembuRitual [Ithaca, N.Y. 1967], p. 38). In Timeand theOther,
ch. 2 ("Our Time, Their Time, No Time: Coevalness Denied," pp. 37-69), Fabian
adds a thirdproblem,the illusion of coevalness dialogism implies,whereas the writing
of ethnographies undermines that coevalness.

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions
320 NEW LITERARY HISTORY

26 Genette's term for bits of narrative that cannot be placed chronologically.See

G6rard Genette, NarrativeDiscourse:An Essay in Method,tr. Jane E. Lewin (Ithaca,
N.Y., 1980), pp. 40, 84.
27 Lejeune, "Autobiographyin the Third Person," p. 44.
28 See Louis Marin, "The Iconic Text and the Theory of Enunciation: Luca
Signorelli at Loreto (Circa 1479-1484)," tr. Lionel Duisit, New LiteraryHistory,14
(1983), 553-96, and his "Towards a Theory of Reading in the Visual Arts: Poussin's
The Arcadian Shepherds,"in Calligram:Essaysin the New Art History fromFrance, ed.
Norman Bryson (Cambridge, 1988), pp. 63-90.
29 For a more detailed critique of the analogy, see my Reading "Rembrandt," pp.
270-72.
30 Gayatri ChakravortySpivak, "Overdeterminationsof Imperialism: David Och-
terlonyand the Rance of Sirmoor,"Europe and Its Others,1 (1985), 131, quoted in
John Barrell, The Infectionof Thomasde Quincy:A Psychopathology of theEmpire(New
Haven, 1991), p. 10.
I am grateful to Norman Bryson, Robert Caserio, Dominick LaCapra, and Ellen
Spolsky for criticalremarkson an earlier version of this paper.

This content downloaded by the authorized user from 192.168.82.204 on Mon, 3 Dec 2012 13:13:14 PM
All use subject to JSTOR Terms and Conditions