PAUL RICOEUR

CAN FICTIONAL NARRATIVES BE TRUE?

The purpose of this essay will be to grasp as a whole the narrative kinds of
discourse ranging from the most "fictional," such as tales, romances, dramas,
novels, and movies, to the most "empirical," including histories, biographies,
and autobiographies. 1 To put this problem in terms that Wittgenstein has
made popular among philosophers, I want to attempt to grasp all these
narrative modes as a unique "language-game." And following Wittgenstein's
example, this means to refer it to some "form of life." 2
That fictional narratives and historical narratives display a certain structural
unity and, in that sense, constitute a unique language-game has been discussed
elsewhere. 3 That the speaking of this language is a part of an activity, or
a form of life, has still to be shown and this will be the central and unique
topic of this essay. For as long as this referential dimension of narrative
discourse has not been acknowledged, the structural unity of the narrative
genre remains problematic, contingent, and, at best, merely factual. Only
the common contribution of the various kinds of narrative discourse to the
shaping of what Wittgenstein calls "life" can provide some kind of necessity
for this factual unity of structure. But what life? Or what aspect of life?
The time has come to remind ourselves that the term history in most of
our Indo-European languages has the intriguing ambiguity of meaning both
"what actually happened" and the report of those happenings. 4 This seems to
be more than a contingent encounter of meanings or a deplorable confusion.
Our language, more likely, maintains, preserves, and indicates by means of
this over determination of the words Geschichte, history, and histoire, a
certain mutual belonging together between telling (or writing) history and
being in history, between doing history and in more general terms being
historical. In other words, the form of life which the speaking of narrative
is a part of is our historical condition itself. To show this will be to solve
the problem of the referential dimension of narrative discourse taken as a
whole.
But this task is not so easy as it may seem. At least three problems are
implied here.
First, to say that history, as history writing, refers to history as our his-
torical condition, seems to limit the referential claim of narrative discourse
3
A-T. Tymieniecka (ed_), Analecta Husserliana, Vol. XIV, 3-19.
Copyright © 1983 by D. Reidel Publishing Company.
4 PAUL RICOEUR

to only that class of narratives which Scholes and Kellogg call "empirical"
narrative, which in practice means some form of history. In order to give to
the referential dimension of narrative the same amplitude as characterizes
the structural component common to both fictional and empirical narrative,
it must be shown that all narratives have, in some sense, a referential claim.
The very distinction between fictional and empirical has therefore to be
questioned in such a way that history may appear to be more jictioruzl than
a merely positivistic conception of history would be prepared to assume and
that fictions in general and narrative fictions in particular appear to be more
mimetic than the same positivistic thought would also allow. So our first
task will be to make sense of the conjunction between fiction and mimesis
in narrative discourse, despite our tendency to deny any role to fiction in
history and to mimesis in tales, dramas, and novels.
Our second task will be to give an independent account of what we mean
by being historical. I speak of an independent account because if we simply
say that all narratives are about our historical condition and that our historical
condition is a "form oflife" which narrative discourse is about, the argument
is clearly circular. A hermeneutic of historicity therefore must rely on its own
categories in order to delineate the ontological structures of the region of
objects which narratives aim at or refer to. Note that this foundational task
does not belong to a mere epistemology or grammar of narrative discourse.
Then, and only then, can it be shown in which different ways the ref-
erential claims of fictional and historical narratives are appropriate to the
ontological structure of historicity. In other words, the fmal problem is to
understand how our fundamental historicity is brought to language by the
convergence of the different modes of narrative discourse and with respect to
their different, yet complementary referential claims.

1. FICTION AND MIMESIS IN NARRATIVE DISCOURSE

It is in no way my intention to cancel or to obscure the obvious differences

between history and the whole set of fictional narratives in terms of their
truth·daims. There is a level of analysis and of argumentation for which
a conventional concept of truth, defined in terms of empirical verification
and falsification, is perfectly valid. Just as Carl Hempel, Ernest Nagel, and
others have forcefully demonstrated that historical explanations do not
fundamentally differ from explanations in the natural sciences, historical
evidences do not entail another concept of truth than the one which is
assumed, say, by physics. Documents and archives are the "sources" of
 CAN FICTIONAL NARRATIVES BE TRUE? 5

evidence for historical inquiry. Fictional narratives, on the other hand, ignore
the burden of providing evidences of that kind. Indeed, I would like to
propose that imagination is the archive for fictional narration. 5 I mean
imagination understood in terms of its temporal complexity as the depository
of oral and written traditions. 6 But even imagination ignores the hard labor
of fighting with documents, and even of establishing them as a function of
the questions addressed to them. In this sense, imagination has no "facts" to
deal with.
This sheer opposition at the level of the epistemology of the social sciences
is not in dispute. It will not be abolished by the further steps of our investiga-
tion. It will even be maintained as a significant contrast in the last stage of
our presentation when we consider what I earlier called the convergence of
the different modes of narrative discourse in the bringing to language of our
fundamental historicity.
It is therefore for a further inquiry that a parallelism of a new kind may
be found between the referential claims of both history and fiction. By a
referential claim I mean the claim to be about something extra-linguistic. I am
thus giving reference the same amplitude that Nelson Goodman gives this
notion in his Languages of Art when he extends it to every kind of symbol,
whether verbal or nonverbal, and to literal and metaphorical applications of
predication, in both it denotative direction (from the symbol to the thing)
and its exemplifying direction (from the sample to the symbol). It is within
this new framework of a general semiotics or symbolics 7 that a certain paral-
lelism between referential claims may be discerned and analyzed, regardless
of the epistemological assymetry in terms of truth-claims. (The question
whether another concept of truth also has to be framed will belong to the
hermeneutical inquiry of our second part.)
Indeed, the weight of the argument lies mainly on the fictional side of
narrative discourse. Yet a full recognition of the referential dimension of
fictional narratives will be made more plausible if the fictional component of
history writing has also been previously acknowledged.
Let us begin with this preparatory step.
It is not foreign to the general trend of the epistemology of history to
underscore the character of "imaginative reconstruction" in the writing of
history. This expression comes from Collingwood, even though he insisted on
the task of "reenactment" in historical knowledge. Thus, while the whole
neo-Kantian school of the philosophy of history, as presented for example by
Raymond Aron in his The German Critical Philosophy of History, 8 tends to
enlarge the gulf between what actually happened and what we historically
6 PAUL RICOEUR

know, it is mainly by the means of a kind of transfer from the theory of

narratives in literary criticism to history considered as a literary artifact 9 that
history writing has begun to be reassessed along the categories of what may
be called semiotics, symbolics, or poetics. In this regard, the influence of
Northrop Fyre's The Anatomy of Criticism and Kenneth Burke's A Grammar
of Motives has been overwhelmingly decisive, especially when taken in
conjunction with such works as Gombrich's critique of the visual arts in Art
and Illusion and Erich Auerbach's great Mimesis: The Representation of
Reality in Western Literature. 1O From these works has emerged a general
concept of the "fictive" representation of "reality" whose scope is broad
enough to be applied also to history writing as well as to fiction.
Hayden White calls his recent investigation into the historical imagination
in nineteenth-century Europe, Metahistory. But he also calls the identification
of the explanatory procedures which history has in common with other
literary expressions of story-telling, Poetics. He borrows from Northrop Frye
the notion of "explanation by emplotment" and the distinction between
the four basic kinds of "plot": romance, tragedy, comedy, and satire. The
historian, according to this point of view, does not merely tell a story. He
makes an entire set of events, considered as a completed whole, into a story. 11
Then White adds to this level of conceptualization another level which he
calls "explanation by formal argument." It is on this level that the historian
seeks to explicate "the point of it all" or "what it adds up to." So to this
level belong the putative laws of the Hempelian model which we discussed
in our first lecture, as well as the putative laws of a dialectical kind such as
the Marxist laws of development applied to global social formations. Next,
following Stephen Pepper's analysis in his World Hypotheses, White differen-
tiates "four paradigms of the form that a historical explanation, considered
as a discursive argument, may be conceived to take: formist, organicist,
mechanistic, and contextualist." 12 Thus the causal (and statistical) model
of explanation is not just one among several paradigms of explanation by
argument, but it belongs to a level of explanation which as a whole makes
sense as intermediary between explanation by emplotment and "explanation
by ideological implication." According to White,

the ideological dimension of a historical account reflects the ethical element in the
historian's assumption of a particular position on the nature of historical knowledge and
the implications that can be drawn from the study of past events for the understanding
of present ones. 13

To the extent that the great historian cannot help taking a position in the
 CAN FICTIONAL NARRATIVES BE TRUE? 7

present world of social praxis, the implicit "set of prescriptions" which rule
his "metapolitical position" may be approached on the basis of Mannheims'
categorization of ideologies: anarchism, conservatism, radicalism, and liber-
alism.14
I have evoked Hayden White's Metahistory as one example of the kind
of inquiry which can be done about historical writing. Two kinds of mis-
understanding must be avoided, however. The first one would be to confme
the above described procedures (or similar ones) to a mere didactic proce-
dure extrinsic to historical inquiry proper. 15 The second one would be to
conceive of this "fictive" representation of reality as exclusive of the rules of
"evidence" that history shares with the other sciences - although admittedly
in its own peculiar way as documentary evidence. Contrary to the first
contention I would hold that the three of four levels of conceptualization
relevant to a "poetics of history" are intrinsic to historical understanding
per se. Through such procedures, events are truly made into a story. But
contrary to the second contention, I should want to stress that as "fictive" as
the historical text may be, its claim is to be a representation of reality. And
its way of asserting this claim is to support it by the verificationist procedures
proper to history as a science. In other words, history is both a literary artifact
and a representation of reality. It is a literary artifact to the extent that, like
all literary texts, it tends to assume the status of a self-contained system of
symbols. 16 It is a representation of reality to the extent that the world that
it depicts - which is the "work's world" - is assumed to stand for some
actual occurrences in the "real" world. 17
That "fiction" and "representation of reality" are not antithetical terms
must now be shown directly with regard to fictional narratives.
The paradigm for this conjunction may already be found in Aristotle's
Poetics. In dealing with tragedy - that is, with poiesis par excellence - he
has no difficulty in saying, on the one hand, that the essence of poiesis is the
mythos of the tragic poem, while, on the other hand, the scope of poiesis is
the mimesis of human action.
The term mythos, by which Aristotle designates the narrative fiction
proper to tragedy, is intended to underscore, first, the fact that it is a kind of
discourse - and, in fact, one of the oldest meanings of mythos is "saying";
second, the fact that it is a fable, a work of fantasy; and above all, that it has
the structure of a plot. This third meaning is the most important. Poiesis is
just that craftsmanship thanks to which the poet "makes" an intelligible
story out of some traditional myth, chronicle, or account. And in this sense,
the myth as of tragedy is like a painting which construes on the basis of a
8 PAUL RICOEUR

system of pictorial signs - of lines and colors - the icon which we learn to
read according to the conventional rules which generated it. The mythos of
drama is therefore to time what the icon of painting is to space. They are
temporal and spatial "messages" generated by a basic "grammar" ruling the
articulation and the combination of the pertinent traits which constitute the
"alphabet" of the playwright or the painter .18
As to the second term of the pair mythos-mimesis, a recurring misunder-
standing keeps obscuring its original meaning. That is, we tend to translate
mimesis by imitation with the sense of copying some already existing model.
But Aristotle had in view a quite different kind of "imitation," a creative
imitation. 19 First, "imitation" is the concept which distinguishes human
crafts and arts from nature. In this sense, it separates more than it connects.
Secondly, there is mimesis only where there is a doing. Mimesis therefore is
homogeneous to poiesis as the construing of a plot. Finally, what mimesis
imitates is "not the actuality of events, but their logical structure, their
significance."2o Mimesis is so little reduplication of reality that tragedy
"intends to represent men better (beltiones) than they actually are" (Poetics
1448a 17-18). Tragic mimesis reenacts reality - that is, human action, but
according to its essential and magnified traits. Mimesis, then, is a kind of
metaphor of reality. And like metaphor, it shows by "signifying the thing
in actuality" (Poetics 1448a 24). In this sense, the mimesis of tragedy is
the counterpart of a similar effect of painting: by abbreviating, condensing,
and combining the pictorial signs of its "alphabet" the painter enhances the
shapes and the colors of the universe. This kind of "iconic augmentation"21
provides an illuminating analogy for the magnifying effect of the tragic
mythos.
I should like now to take this conjunction between mythos and mimesis
in Aristotle's Poetics as the paradigm for the kind of referential claim which
seems to me to be relevant for fictions in general. The recognition of this
referential claim is often prevented by the prejudices which prevail in the
field of our ordinary theory of the imagination. These stub born prejudices
tend to make the term image synonymous with that of a copy or a replica of
a previously given reality. And, in fact, these prejudices may be traced back
to common sense. Ordinary language tends to impose as the paradigmatic
case of image either the physical replica of some absent thing which could be
shown or perceived elsewhere, or the mental representation of it. Philosophies
of ordinary language, such as that of Gilbert Ryle (as well, in this instance,
Jean-Paul Sartre!), follow this ordinary usage and tend therefore to take
for granted the defmition of the image as the intuitive representation of
 CAN FICTIONAL NARRATIVES BE TRUE? 9

something in absentia. The question of the referent of the image then raises
no specific problem since by hypothesis the same thing may be perceived in
praesentia or imagined in absentia. The difference between the image and
perception concerns only modes of givenness.
Philosophers used to reinforce this notion of the image conceived of as
a reduplication of reality by deriving simple images from corresponding
impressions and by deriving fictions from these simple images by means of
their being associated into complex ideas. Fictions then were held to be
merely new combinations of components borrowed from previous experience,
as may be demonstrated if we recall those two favorite examples of such
philosophies: the centaur and the chimera. But what is overlooked in such
an approach is that perception is already selective and interpretative in its
activity. And one also tends in this way to forget that the enigma of fiction
resides precisely in the newness which occurs in the ordering of appearances.
This reluctance of philosophers to meet the challenge of novelty explains
why the notion of a productive reference must appear at first sight as an
unbearable paradox. As we just said, the image as a copy escapes this paradox
to the extent that the image and corresponding perception have the same
referent, the thing out there, and they only differ according to the modes
of givenness of one and the same thing. The lack of reference in fiction to
a previous given raises the truly critical problem, however, for a theory of
imagination which brings to the forefront the difference between reproductive
and productive imagination. Why? Because fiction alone raises the problem
of the unreal as distinct from the mere absence of the real. Whereas the
nothingness of absence concerns only modes of givenness, the nothingness of
unreality concerns the referent of fiction.
But this paradox of unreality is only half of the enigma. By saying that
fiction is without an original given that it would reduplicate in absentia,
we deny only one way of referring to reality, that of the representative or
reproductive mode of reference. But at the same time and by the same token,
the way is paved for another mode of reference. Because fictions do not
refer in a reproductive manner to reality as already given, they do refer in a
productive manner to reality as prescribed by them.
The field of application of such a theory of fiction and of productive
reference is immense. It has the same amplitude as a theory of symbols, under-
stood in the sense of Ernst Cassirer and Nelson Goodman. With the latter, we
may equate productive reference with "reality remade," the title of the first
chapter of his book, Languages of Art. Symbolic systems make and remake
reality. This is the case with aesthetic icons, but also with epistemological
10 PAUL RICOEUR

models, as well as with political utopias. All of them are cognitive in the sense
that they make reality appear as it does. All of them have this organizing
power because they have a signitive dimension, because they are brought
forth by work and craftsmanship, and because they generate novel grids for
reading experience. Note that these are the three traits we already encoun-
tered in Aristotle's concept of mythos: saying, making, and emploting.
Furthermore, the signitive trait of fiction is overlooked in typical accounts
of the image in terms of residual impressions. 22 Its reliance on work is ignored
in accounts of images in terms of mental occurrences lacking confrontation
with an external medium; 23 and the novelty of fiction is denied when its
structural texture is not assigned to the configurational operation of the
reflective judgment.24
But when these three basic features of fictions are understood on their
own terms, the notion of productive reference loses its paradoxical appear-
ance. It means that fictions "reorganize the world in terms of works and
works in terms of the word."25 Or to use a vocabulary which belongs rather
to the epistemology of models than to the theory of aesthetics, fictions
redescribe what conventional language has already described. 26 By connecting
fiction and redescription, we merely give its full extension to the connection
Aristotle established in his Poetics between mythos and mimesis.
The time has come to apply this general theory of fiction to narrative
fiction. It is all the more easy to do so in that our initial example was a
particular case of narrative fiction. We have only to generalize the Aristotelian
theory of drama to all narrative fictions on the basis of what we have said of
fictions in general.
Like all poetical work, narrative fictions proceed from an epoche of the
ordinary world of human action and of the descriptions of this ordinary
world in ordinary discourse. Description has to be suspended in order that
redescription may occur. This negative relation to reality has often been
emphasized and sometimes even over-emphasized by literary critics eager to
provide literature with an autonomous status of its own. Hence the tendency
to oppose the "poetic" function of a given message to its "referential"
function. To the extent that a poem would be for the sake of its own inner
verbal texture,it would not be about the world. But we saw that the reference
of fiction to the unreal is only the negative counterpart of its productive
reference. Or, to put it in other terms, the suppression of a first order refer-
ence - which we have called conventionally the "description" of the world
- is the condition of the possibility of a second order reference which we are
here calling the redescription of the world. A literary work, it seems to me, is
 CAN FICTIONAL NARRATIVES BE TRUE? 11

not a work without reference, but a work with a split-reference, i.e., a work
whose ultimate reference has as its condition a suspension of the referential
claim of conventionallanguage. 27
So we are now prepared to formulate the central thesis of this essay: both
history and fiction refer to human action, although they do so on the basis of
two different referential claims. 28 Only history may articulate its referential
claim in compliance with rules of evidence common to the whole body
of science. In the conventional sense attached to the term ''truth'' by the
acquaintance with this body of science, only historical knowledge may
enunciate its referential claim as a "truth"-claim. But the very meaning of
this truth-claim is itself measured by the limiting network which rules the
conventional descriptions of the world. This is why fictional narratives may
assert a referential claim of another kind, appropriate to the split reference
of poetic discourse. This referential claim is nothing other than the claim to
redescribe reality according to the symbolic structures of the fiction. And
the question, then, is whether in another sense of the words true and truth,
history and fiction may be said to be equally "true," although in ways as
different as their referential claims are themselves different.

2. TOWARD A HERMENEUTIC OF HISTORICITY

To answer the question of the ground of reference for both history and
fiction we must return to the rich ambiguity of the very word history which
we pointed to earlier. Geschichte, history, and histoire - we said - mean
both what really happened and our accounts of those events. 29 What is at
stake in this new phase of our inquiry is the right to continue to use the word
history in both ways once we have exposed its ambiguity. We can do so only
if we can make sense of the mutual belonging together of the epistemological
and the ontological sense of the word history. This mutual belonging together
in tum entails the concept of truth we are looking for and which should
encompass both the referential claim of fictional narratives and the referential
claim of historical narratives (as "true" stories in the epistemological sense of
"true"). In other words, our problem is to show how the speaking of our
narrative language is part of a form of life. 30
This investigation exceeds the possibilities of a poetics of historical imagi-
nation and requires a hermeneutic of historicity.3! This hermeneutic of
historicity, in tum, must begin, at least in my opinion, with Kant, although
it cannot simply remain with him.
It must begin with Kant because he is the first modem thinker to have
12 PAUL RICOEUR

linked the problematic of imagination with that of time. And this is what
is fundamentally required by an investigation into the basic relationship
between historical imagination and human temporality. Without entering into
the technicalities of Kant's work, it may suffice to remind ourselves here that
in the Transcendental Aesthetic the question of time was more or less dealt
with without the narrow framework of a parallelism between space and time,
although the "inner sense" (to which time is ascribed and with which it is
nearly identified) is said to encompass all experience, be it outer or inner. But
this inner sense would remain the mere form of a manifold if it were not
gathered, organized, and structured by a synthetic activity. And this activity
is ascribed to the imagination under the well-known title of the "synthesis of
reproduction in the imagination." Even more decisive for our enterprise is the
discovery that this synthesis of reproduction would not be possible without
the spontaneous activity of a productive imagination which makes possible
the reproducibility of phenomena. This is basic for the understanding of
the way in which the manifold of our life-stories is "run through" and "held
together" by our historical imagination which makes this manifold into
a story. In tum, the way in which this historical imagination is part of a
form of life make less cryptic the connection made by Kant between the
productive imagination and the reproducibility of appearances. It is as though
the spontaneity of story-telling works as a retrieval of the contingent and
passive links of our temporal experience.
This comparison with Kant's philosophy may still be pushed two steps
further before we leave it behind. In the first Critique, you may recall, the
theory of imagination reaches its climax with the doctrine of the schematism.
There imagination is less a faculty of contemplating fixed and determinate
images than a procedure, a method for construing such images for the sake of
conceptual activity. And in the same way, it has been suggested at the end of
our second lecture that the whole tradition of narratives, to the extent that it
presents stable and recurrent patterns, constitutes the schematism through
which the logic of possible narratives may be applied to the nearly endless
field of stories, tales, dramas, novels, etc. and through them to our temporal
experience itself.32
Finally, our last and most decisive borrowing from Kant is from his
theory of reflective judgment in the Critique of Judgment and from the
new description of imagination which this theory entails. Louis O. Mink
is, to my knowledge, the first analytic philosopher to underscore the kin-
ship between the configurational act of historical knowledge and Kant's
reflective judgment: through narratives, we reflect upon events by retelling
 CAN FICTIONAL NARRATIVES BE TRUE? 13

and rewriting them. We not only grasp them together, according to the
synthesis of the imagination from the first Critique, but we make them into
organized wholes submitted to teleological apprehension. Imagination, at
this level, is that free play of the faculties which satisfies both the freedom
of invention and a strange kind of order that he calls legality without laws
under the guidance of the finality without purpose of the reflective judg-
ment. What better illustration of this ruled creativity may be found than the
spectacle of the "imaginative variations" displayed by the infinite variety
of the stories told by mankind and its spontaneous submission to recurrent
patterns?
But if it is in Kant that we fmd the first attempt to connect imagination
and time, the narrative intelligibility displayed by the activity of story-telling
exceeds the resources of his philosophy. And it exceeds it in two ways. First,
the activity of narration requires an intersubjective dimension which is lacking
in both the first and the third Critiques (to say nothing of the second which is
not at stake here). Second, the connection between historical knowledge and
historicity can no longer be put in the terms of a philosophy understood as
a Critique.
It is important to say that story-telling displays its imaginative skill at the
level of a human experience which is already "communalized." Plots, char-
acters, thematic elements, etc. are forms of a life which is really a common
life. In this respect, autobiographies, memoires, and confessions are only
subsections of a narrative art which as a whole describes and redescribes
human action in terms of interactions. 33 Therefore a philosophy which links
imagination and time at the level of some "inner sense" is inadequate to the
intersubjective dimension of historical imagination. Kantian philosophy
cannot make this move because the ultimate ground of synthesis for it is the
unity of apperception which is identified with the abiding and timeless ego of
the "I think."
A hermeneutic of historicity must assume another founding ground. It
must assume as its first presupposition that my temporal field is primordially
"paired," to use the language of Edmund Hussed, with other temporal
fields.34 This Paarung implies in turn that the unfolding of my flow of
consciousness is accompanied by that of my contemporaries in a sense quite
different from the sense in which the Kantian "I think" can accompany
all my representations. This "pairing" is not even exhausted by the concept
of contemporaneity. Historical imagination refers to a broader temporal
field within which my personal history is related in a threefold way to the
temporalities of my contemporaries, my predecessors, and my successors, to
14 PAUL RICOEUR

use the language of Alfred SchutZ. 35 Such must be the primordial constitution
which makes possible historical judgment and historical imagination.
The gap between this hermeneutic of historicity and Kantian philosophy
becomes larger when we consider the conditions of intelligibility which rule
the historical field of experience. As close as these conditions may be to the
teleological judgment as formulated by Kant, they have to be articulated in a
new conceptual framework, a framework that should benefit equally from
contributions from both existential phenomenology and the philosophy of
language analysis.
The specific categories of the limiting framework of historical under-
standing would include such notions as the following: human agents who
begin events of which they are the authors; these agents' interpretations of
their actions in terms of motives; the influence of one agent on another who
takes the meaning of the first agent's action into account; the regulation
of such projects by norms and the regulation of norms by institutions; the
founding of such institutions and their sedimentation; and the continuation,
breaking off, or renewal of contents so transmitted. In short, historical
transmission needs to be thought of differently than as succession as it is
conceived of in the natural sciences.
At this point of our investigation into the structures of historicity which
ground the referential claims of both history and fiction we must acknowledge
that we have not only expanded upon the philosophy of Kant by adding to
it an intersubjective dimension which it lacks, but that we have implicitly
broken with philosophizing conceived of as a critique in the Kantian sense of
this term.
The intersubjective field of temporal experience which all our stories
refer to can no longer be treated as an object standing "over against us," as
a Gegen-stand. We ourselves, in turn, are no longer the subject "for" whom
there are such temporal phenomena. The subject-object relationship itself
is undermined by the very progression of reflective investigation from the
theory of "inner sense" to that of intersubjective historicity. We are part of
this field as storytellers and as story-followers, as historians and as novelists.
We belong-to history before telling stories or writing histories. The historicity
proper to story-telling and history-writing therefore is encompassed within
the reality of history. Such is the fundamental relationship that obtains
between the two meanings of the word history. As Hans-Georg Gadamer
says, "a proper hermeneutics would have to demonstrate the effectivity of
history within understanding itself. I shall refer to this as 'effective-history'
(Wirkungsgeschichte). Understanding is essentially an effective-historical
relation" (Troth and Method, p. 267).
 CAN FICTIONAL NARRATIVES BE TRUE? 15

3. HISTORICITY AND THE TWO NARRATIVE MODES

The previous inquiry into the intelligible constitution of historicity allows us

to return to our initial problem of the complementarity between fictional and
empirical narratives. My contention in this regard is that this complementarity
is not only grounded in this intelligible constitution but is required by it. We
need both the empirical and the fictional to bring to language our historical
situation.
In order to show this, I want to explore the ways in which each narrative
mode shares something of the intentionality of the other. My contention,
then, is that it is in this exchange between history and fiction and between
their opposite referential claims that our historicity is brought to language.
At first glance this contention looks like a sheer paradox because it runs
against two well-established conventions. On the one hand, we tend to ascribe
to history a mere scientific function, in the sense of a pure concern for
"objectivity." On the other hand, we tend to ascribe to fiction a merely
"subjective" function in the sense of a play of our imagination for the sake
of entertainment. But this opposition between earnest inquiry and playful
entertainment is only true for a first order investigation which isolates meth-
odological procedures. A second order investigation would address itself to
the interests which rule such methodology. By an interest, I mean exactly
what Kant meant when he raised the question of the different interests of
reason involved in the thesis of a free causality and the antithesis of necessary
causality. An interest is not simply a Psychological factor. It concerns the
goals which orient a cognitive activity.
If we look for the interests which animate historical inquiry, we get a
much more sophisticated answer than if we simply consider its methodology
in abstraction. A conventional concept of objectivity common to natural
and social sciences may satisfy the first order inquiry into methodology,
but this is no longer the case for a second order inquiry into interests. Then
the pure interest for facts seems to be coupled with another more deeply
rooted interest which I should like to call, following Habermas, an interest for
communication. In other words, our ultimate interest when we do history is
to enlarge our sphere of communication. This interest expresses the situation
of the historian as belonging to the field that he studies. He is a member
of his set of objects. Consequently, every procedure of objectivation, of
distanciation, of doubt, of suspicion, and in short, everything that makes
history a form of inquiry is a part of this interest for communication. How?
In at least two ways.
First, the historian tries to keep from the past what deserves not to be
16 PAUL RICOEUR

forgotten, what is, in the strict sense of the word, memorable. And what is
most worthy of being kept in our memories are the values which ruled the
individual actions, the life of the institutions, and the social struggles of the
past. Thanks to the objective work of the historian, these values are added to
the common treasure of mankind.
But this way of resuming the forgotten past requires as its counterpart
a capacity of suspending one's own condition, an aptitude to practice the
epoche, the bracketing of one's own desires. And thanks to this epoche the
alterity ofthe other is preserved in its difference. 36 Hence a dialectic between
what is foreign and what is familiar, between what is remote and what is near,
is at work in what I am calling the interest for communication that rules even
the "objective" attitude of the historian. And this dialectic is precisely what
brings history close to fiction - because the recognition of the values of the
past in their difference opens the real to the possible. History, too, in this
respect, belongs to the logic of narrative possibilities. It does so, however,
not through fiction, but precisely by means of its "true" stories. The "true"
stories of the past expose the potentialities of the present. Croce used to say
that there is no history but the history of the present. This is a paradox of
course. But it does say something. It says that history is, in a sense, the field
of the "imaginative variations" that surround the present and the actual as it
is taken for granted in daily life.
Briefly stated, such is the way in which history, precisely because it is
"objective," shares something of the fictional world. The reverse is no less
true: fictional narratives share in some way the realistic intention of history.
All that we said about the mimetic dimension of fiction allows us now to
conclude that, thanks to its mimetic intention, the world of fiction ultimately
redirects us to the core of the actual world of action. Aristotle, again with
reference to Greek tragedy, said paradoxically that "poetry is more philo-
sophical . _. than history" (Poetics 1451b 5-6). What he meant is that
because history is tied to the contingent it misses the essential, whereas
poetry, not being the slave of the real event, can address itself directly to the
universal, i.e., to what a certain kind of person would likely or necessarily say
or do.
Therefore, to conclude, could we not say that by opening us to the dif-
ferent, history opens us to the possible, while fiction, by opening us to the
unreal, brings us back to the essential?

Universities of Paris and Chicago

 CAN FICTIONAL NARRATIVES BE TRUE? 17

NOTES

1 I retain the terminology used by Scholes and Kellogg in The Nature of Narrative
(Oxford: Oxford University Press, 1968) for didactic purposes. I shall later qualify the
dichotomy between fictional and empirical, however.
2 The "speaking of language is part of an activity, or a form of life," Philosophical
Investigations #23. "To imagine a language means to imagine a form of life," ibid.,
#19.
3 Paul Ricoeur, 'The Narrative Function' in Hermeneutics and the Human Sciences, ed.
and transl. by John B. Thompson (Cambridge: Cambridge University Press and Paris
Maison des Sciences de l'Homme, 1981),274-96.
4 H. Marrou (paris: Editions du Seuil, 1954), p. 36, has noted that all the attempts
to reserve the German Geschichte for the first meaning and to ascribe the second to
Histoire have failed. Something deeply rooted in the wisdom of our languages resists this
clear-cut dichotomy.
5 There is between the use of reflective judgment in fictional narrative and its use in
historical narrative the same difference as in Kant's third Critique between its use in an
aesthetic judgment and in the teleological judgment about nature. The difference lies in
the Darstellung (exhibitio), the "presentation" of the finality of forms. "This may take
place either through our own imagination, as in art when we realize a preconceived
concept of an object which is a purpose of ours, or through nature in its technique (as
in organized bodies), when we supply to it our concept of its purpose in order, to judge
of its products." The Critique of Judgment, 'Introduction,' #VIII (Barnard trans.,
p.29).
6 Regarding the "oral heritage of written tradition," see Scholes and Kellogg, The
Nature of Narrative, chap. 2, pp. 17-56, and Albert Lord, The Singer of Tales (Cam-
bridge: Harvard University Press, 1960).
7 The subtitle of Goodman's book Languages of Art (Indianapolis, The Bobbs-Merrill
Co., 1968) isAn Approach to a Theory of Symbols.
8 Raymond Aron, La Philosophie critique de I'Histoire, Essai sur une tMorie allemande
de l'Histoire (paris: Vrin, 4ieme ed. 1969).
9 See, for example, Hayden White, 'The Historical Text as Literary Artifact,' Clio 3
(1974): 277-303.
10 This complex filiation is acknowledged by Hayden White in his 'Introduction: The
Poetics of History' to Metahistory: The Historical Imagination in Nineteenth Century
Europe (Baltimore: The Johns Hopkins University Press, 1973).
11 "For example, Michelet cast all his histories in the Romantic mode, Ranke cast his
in the comic mode, Tocqueville used the tragic mode, and Burckhardt used satire." Ibid.,
p.8.
12 Ibid., p. 13.
13 Ibid., p. 22.
14 White adds to these three levels of explanation a further level which concerns the
"historiographical styles," i.e., the modes which express the elective affinities which
rule the combinations between modes of emplotment, modes of argument, modes of
ideological implication. This stylistic level may be characterized on the basis of the
classical theory of tropes (metaphor, metonymy, synedoche, and irony), which allows
the identification of the kinds of figurative language used. It should be noticed, however,
18 PAUL RICOEUR

that the author does not reduce his poetics to a tropology and he does not treat his
tropology as a substitute for any of his three levels of explanation.
15 See Maurice Mandelbaum, 'A Note on History as Narrative,' History and Theory 6
(1967): 413-419, p.414.
16 See Northrop Frye, The Anatomy of Criticism (Princeton, New Jersey: Princeton
University Press, 1957), the fust essay: 'Historical Criticism: Theory of Modes'.
17 See Beardsley, Aesthetics, Bk V, #15; Hans-Georg Gadamer, Truth and Method
(New York: The Seabury Press, 1975), pp. 410-429, on "the thing of the text" (in the
sense of its "cause" or "issue").
18 See E. H. Gombrich, Art and Illusion (London/New York: Phaidon Books, 1960),
chap. I.
19 See R. McKeon, 'Literary Criticism and the Concept of Imitation in Antiquity,'
in R. S. Crane (ed.), Critics and Criticism (Chicago: The University of Chicago Press,
1952,1970 5 ).
20 Leon Golden and V. B. Hardison,Aristotle's Poetics, A Translation and Commentary
for Students of Literature (Englewood Cliffs, NJ.: Prentice-Hall, 1958), pp. 68-69, and
the Epilog: 'On Aristotelian Imitation', pp. 281-296.
21 See F. Dagognet, Ecriture et /conographie (paris: Vrin, 1973).
22 It should be the function of a theory of poetic images to detach the problematic of
the image from the bondage of perception and to connect the emergence of the image to
that of language; "The poetic image brings us to the origin of speaking being." Gaston
Bachelard, The Poetics of Space, 'Introduction.'
23 Plato's condemnation of writing in the Phaedrus entails that of painting and of all
other kinds of "images" as well to the extent that they all rely on "external marks"
instead of pure reminiscence. This critique remains implicit in all descriptions of images
as a mere "shadow" of what is. But painting, as we said above, offers an alternative:
because its "external marks" are wrought, the icon may augment reality.
24 In the sphere of language, the metaphorical process is a good example of semantic
innovation. That live metaphors are untranslatable without semantic loss proves that
there may be something truly new in the order of meanings. See my La Metaphore vive,
Etude III.
25 Nelson Goodman, op. cit., p. 241.
26 See Mary Hesse, 'The Explanatory Function of Metaphor,' in her Models and Analo·
gies in Science (Notre Dame University Press, 1966, 1970), pp.157-177.
27 I develop the theory of double reference at greater length in La Metaphore vive,
Etude VII. The notion of a split reference is borrowed from Roman Jakobson. This
notion qualifies his fust statement that the poetic and the referential function are
antithetical. See his 'Two Aspects of Language and Two Types of Aphasia Disturbances'
in Fundamentals of Language (The Hague: Mouton, 1956), quoted in MV, p. 282.
28 My use of the term reference in relation to narrative fictions corresponds to the
concept of meaning in Scholes' and Kellogg's The Nature of Narrative: "Meaning, in a
work of narrative art, is a function of the relationship between two worlds: the fictional
world created by the author and the 'real' world, the apprehendable universe," p. 82.
29 The third sense of the word event in our discussion of the anti-narrativist arguments
in 'The Narrative Function' (see Note 3 above) - event as the boundary~oncept of
"what actually happened" - pointed implicitly toward this ontological use of the word
history. Epistemology is connected to ontology through this limit~oncept.
 CAN FICTIONAL NARRATIVES BE TRUE? 19

30 See above, p. 3.
31 As I said in my opening remarks, the argument that history and fiction are about our
being historical and that our being historical is what narratives are all about would be
merely circular if we had no other means to characterize our historical condition.
32 This schematism of the logic of possible narratives makes possible a typology of plots,
characters, and topics such as the one proposed by Northrop Frye in his "archetypal
criticism" in the second and third essays of his Anatomy of Criticism.
33 Scholes and Kellogg clearly show how individual memory is restructured by the
internalization of public stories in the form of "confessions."
34 See E. Husserl, Cartesian Meditations, #5, par. 39.
35 See A. Schutz, The Phenomenology of the Social World (Evanston: Northwestern
University Press, 1967), pp. 139-214.
36 See here again Veyne concerning the "inventory of differences."