CONNERTY, 1990. Historys Many Cunning Passages. Paul Ricoeurs Time and Narrative

Duke University Press and Porter Institute for Poetics and Semiotics are collaborating with JSTOR to digitize, preserve and extend access to Poetics Today.

http://www.jstor.org

Porter Institute for Poetics and Semiotics

History's Many Cunning Passages: Paul Ricoeur's Time and Narrative Author(s): J. P. Connerty Source: Poetics Today, Vol. 11, No. 2, Narratology Revisited I (Summer, 1990), pp. 383-403Published by: Duke University PressStable URL: http://www.jstor.org/stable/1772623Accessed: 12-05-2015 14:06 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

This content downloaded from 139.82.115.34 on Tue, 12 May 2015 14:06:43 UTCAll use subject to JSTOR Terms and Conditions

History's Many Cunning Passages: Paul Ricoeur's Time and Narrative

J. P. Connerty College of Law, London

In the Three Essays on the Theory of Sexuality, Freud criticizes the popular conception of sexual development, which "is beautifully reflected in the poetic fable which tells how the original human beings were cut up into two halves-man and woman-and how they are always striving to unite again in love" (1962 [1942]: 45-46). Freud offers in contrast his own narrative account, which radically revises the conception of infantile sexuality and can better explain the existence of sexual "aberration," but which in doing so raises the question of how any one story can be "better" than another. In The History of Sexuality, Michel Foucault recounts another popular conception of sexuality, in which, "the story goes," the frankness of the seventeenth century is replaced by the repression of the Victorian age, which continues to dominate today (1987: 3-6). These two popular stories can be con- trasted as a partially formalized account of individual sexuality versus a more properly historical account of sexuality in general, but the most striking point is that, unlike Freud, Foucault does not offer an alternative narrative account of his own, instead proposing an inquiry into the reasons why we should ever have believed the popular story at all.

Foucault's apparent rejection of narrative is very significant; of all the human sciences, history is the one that most obviously employs narrative techniques, and historiographers have given a great deal of thought to the role of narrative in history writing. As Peter Gay comments, "Historical narration without analysis is trivial, historical analysis without narration is incomplete" (1975: 189). Poetics Today 11:2 (Summer 1990). Copyright ? 1990 by The Porter Institute for Poetics and Semiotics. ccc 0333-5372/90/$2.50.


384 Poetics Today 1 1:2

The reciprocity of the relationship between the two, which this quo- tation may be taken to imply, is rather an unusual conception for an historian; the relationship is more commonly thought of as supplementary. But as we have seen such reciprocity follows from the argument that a concern with narrative as an object implies a dependence on narrative as a method; and history also is subject to this argument. As Natalie Zemon Davis argues, although scientific historians are taught "to peel away the fictive elements in [their] documents to get at the real facts," this is difficult, if not impossible, to do, because the re- counting of these facts is accompanied by (and not followed by) "shap- ing choices of language, detail, and order" (1988: 3). Thus Davis's account of sixteenth-century letters of remission stresses their fictive element, demonstrating the way in which they were told and retold by different voices before being presented to the king as a finalized narrative account (ibid.: 7-36). Her own account of this process is itself a partially formalized narrative, abstracted from a large number of individual cases; as a result we are led to wonder to what extent her comment that "remission narratives had to persuade [the king and magistrates] and establish a story that would win out over competing versions" can be applied to her own narrative history (ibid.: 15).

As Dominick LaCapra points out, although historians have renewed their inquiry into the role of rhetoric in the objects of their study, they have failed to follow literary criticism and philosophy in making the reflexive and self-critical move that examines the role of rhetoric in their methodology: "Understandably fearful of the involuted, narcissistic ex- tremes of self-reflection, historians have paid scant attention to their own rhetoric and to the role of the rhetorical (including the rhetoric of so-called 'hard' science) in constituting their own discipline" (1985: 16-17). In fact, I would argue, historiographers have paid a good deal of attention to the narrative status of their methodology and the implications of that narrative status, and it is these implications that I want to discuss. From the nineteenth century onwards, along with efforts to make history into a scientific subject, a distinction was drawn between historical analysis and the supplementary narration of that analysis. Only then did the idea that there might be a problem with the concept of historical narrative-that the phrase "narrative history" might be a contradiction in terms-actually arise. The problem is summarized by Hayden White:

To many of those who would transform historical studies into a science, the continued use by historians of a narrative mode of representation is an index of a failure at once methodological and theoretical. A discipline that produces narrative accounts of its subject matter as an end in itself seems theoretically unsound; one that investigates its data in the interest of telling a story about them appears methodologically deficient. (1987: 26)


Connerty * Paul Ricoeur's Time and Narrative 385

A more negative statement of this anxiety comes from Sande Cohen: A culture is reactive when it continues to narrativize itself despite, at any moment, being six minutes away (by missile) from its own non-narrative obliteration. The dissemination of models of "history" promotes cultural subjects who are encouraged to think about non-narrative relations- capitalism, justice, and contradictions-in a narrative manner. Narrating screens the mind from the non-narrative forces of power in the present, insofar as "historical" narration reduces present semantic and pragmatic thought to forms of story, repetition, and model, all of which service cultural redundancy. (1986: 1) One of the things I want to do is to defend narrative thinking against

this very severe judgment. As a result of this anxiety, historiographers tend to divide into narrativist and antinarrativist groups, and I want to try to relate those positions to, on the one hand, historical relativism, and on the other hand, scientific history or historicism, then to argue that Ricoeur attempts to transcend these oppositions altogether.

Historiography is unusual among the human sciences in containing a body of work that is not simply an apologia for narrative but a posi- tive endorsement of it. This attitude is found in the Anglo-American analytical philosophers A. C. Danto (1985 [1968]), W. B. Gallie (1964), Louis O. Mink (1978), and W. H. Dray (1957), who believe that narrative form is particularly well-suited to historical explanation and exposition. The valorization of narrative as a mode of historical explanation follows upon the undermining of the "covering-law model" as an adequate model for historiography. The philosopher C. G. Hempel (1949) had argued that historical explanation could and should be subject to the same kinds of general laws that obtained in the natural sciences, and if, in contrast to the natural sciences, historians made use of "empirically meaningless terms," then their accounts would be "pseudo-explanations" and not "scientific explanations" (1949: 465). So, for example, to explain an event such as the outbreak of the Pelo- ponnesian War, the historian would need to base his explanation in a covering-law, a law that was always and everywhere true, such as, "When one country expands its military power, this will lead to an outbreak of war with its neighbour."

W. H. Dray's Laws and Explanations in History was the most radical critique of this position; Dray argued that the "uniqueness" of every historical event negates the functioning of "laws" in history. Even if historians employ "classificatory terms" to discuss, say, the Russian Revo- lution as a "revolution" in the abstract, they do so only to highlight the difference between this (unique) revolution and other (unique) revolutions (1957: 44). However, the "uniqueness" of the historical event implies a relativistic position in which historical differences are em- phasized at the cost of generalization and, consequently, at the cost of


386 Poetics Today 1 1 :2

explanation. The problem can be posed as follows: How can an abstract covering-law be related to an individual historical event (as rule to example of rule), without either making the law context-specific (and thus not an abstract law) or making the event an abstract universal (and therefore not an individual event)? And if it is not possible to talk in general terms about events, how do we avoid historical relativism and retain for history the idea of explanation without objective laws?

Narrative is offered as an alternative logic. W. B. Gallie writes:

What is new in my account of historical understanding is the emphasis that I have put on the idea of narrative. I have tried to analyse what it means to follow a narrative and have argued that whatever understanding and whatever explanations a work of history contains must be assessed in relation to the narrative forms which arise and whose development they subserve. (1964: 9)

The difficulty in grasping such a concept is a result of the bias in West- ern philosophical tradition towards theoretical knowledge, "knowledge admitting of universal statements and of being set out in systematic form" (ibid.: 21). Historical knowledge is not of this type, but neither is it wholly subjective or pragmatic knowledge. The particular task that narrative can uniquely perform for historical explanation is described by Danto: "The role of narratives in history should now be clear. They are used to explain changes, and, most characteristically, large-scale changes taking place, sometimes, over periods of time vast in relationship to single human lives" (1985 [1968]: 255).

Unfortunately, Gallie's and Danto's idea of what a story is and how it functions is a naive one. For example, Danto writes, discussing both historical and fictional narrative unity, ". .. if N is a narrative, then N lacks unity unless (A) N is about the same subject, (B) N adequately explains the change in that subject which is covered by the explanandum, and (C) N contains only so much information as is required by (B) and no more" (ibid.: 251).

None of these criteria of unity applies to postmodernist fiction, for example; and this is significant because, once the centrality of narrative to historical explanation is asserted, then anything that is true of narrative fiction is also true of narrative history. That is to say, if there is something in the nature of history that requires it to be formulated as narrative (this is something that I will return to), then narrative is no longer supplementary to historical analysis but central to it. To get an idea of what Gallie and Danto are not taking into account in their analysis of the story, we should approach the relationship between history and narrative from the other side. J. Hillis Miller (1974) argues that it was traditional for classic realist fiction to attempt to disguise its fictional status and represent itself as history: "By calling a novel



a history its author at one stroke covers over all the implications of gratuitousness, of baseless creativity and lie, involved in the word 'fiction'. At the same time, he affirms for his novel that verisimilitude, that solid basis in pre-existing fact, which is associated with the idea of history" (1974: 457). However, Miller continues, the result is that history becomes associated with the problematics of narrative:

Insofar as a putting in question of its own enterprise has been an intrinsic part of the practice of prose fiction in its modern form from Don Quixote on through Tristram Shandy to John Barth and J. L. Borges, this unravelling has also been a dismantling of the basic metaphor by means of which prose fiction has defined itself, that is, as a certain idea of history. (Ibid.: 462) How can historiography avoid this contamination when it asserts the

likeness of history and narrative? How do history and fiction differ? According to the traditional argument, the difference is that history refers to real events, while fiction refers to imaginary events. So, for example, Louis 0. Mink writes:

History and fiction are alike stories or narratives of events and actions. But for history both the structure of the narrative and its details are representations of past actuality; and the claim to be a true representation is understood by both writer and reader. For fiction, there is no claim to true representation in any particular respect. (1978: 130; see also Gallie 1964: 55) This rather clashes with Miller's remarks concerning realistic fiction.

Both the reference to the "real" events and to the differing "intent" of the historian and novelist put the emphasis on the context of the narrative-the historical context of the historical text. The theoretical status of historical narrative is dependent on the existence of and access to a pre-narrative reality, of which the narrative is a simple copy. But the adequacy of the narrative history as a representation of the historical context can be judged only with reference to that context; knowledge of the context is available only from the text itself-the validity of which is not yet proved. It is the circularity of this argument that will hinder any attempts to create a hierarchy between the historical text and its context, as we shall see.

The most obvious case that refutes the argument exemplified here by Mink is that of Daniel Defoe-ironically, in literary history, himself an origin. Defoe's fictions masquerade as histories, often so successfully that it is simply not known whether some of the "pseudo-histories" are fictional or historical accounts (the latter written in Defoe's capacity as ajournalist). In this case, therefore, these "fictions" may be "histories" -the deciding factor being their relationship with their context, which can only be reconstructed from the text. It can neither be told nor demonstrated from the narrative itself whether it is history or fiction; the context must already be known. This Mink concedes, admitting



a "dilemma" in the concept of "historical narrative": "As historical it claims to represent, through its form, part of the real complexity of the past, but as narrative it is the product of an imaginative construction, which cannot defend its claims to truth by any accepted procedure of argument or authentication" (1978: 145).

The difficulty of differentiating between narrative history and narrative fiction has led some historiographers to deny that there is any real similarity at all. Thus Maurice Mandelbaum (1977) argues that the "narrative" aspect of history is a method of presentation that comes after the process of historical inquiry has finished:

The historian's 'story'-if one chooses to regard it as a story-must emerge from his research and must be assumed to be at every point dependent on it. It is therefore misleading to describe what historians do as if this were comparable to what is most characteristic of the storyteller's art. The basic structure of a story or tale is of the storyteller's own choosing and whatever may be preliminary to his telling that story does not serve to control the act of narration. (1977: 25)

Again the distinction is that the storyteller invents, while the historian inquires-that is, narrative in history is supplementary to a real event or sequence of events, while in narrative fiction there is only the narrative. Thus we see that traditionally the relationship between text and context is not considered problematic. Either the material nature of the text-narrative or otherwise-is unimportant, merely the secondary, discursive manifestation of the facts of the analysis; or, if attention is paid to the narrative/rhetorical status of the text, the distinction between history and fiction is made by easy reference to the "real" event. Similarly, two narrative accounts that differ can be judged by reference to the "real" which they depict.

However, once the centrality of narrative to historiography has been asserted, and the self-reflexive move that LaCapra has called for is made, the problems of the text/context relationship become clear. The most extensive of these critiques is that of Hayden White, who chal- lenges the idea that the historical text is secondary or supplementary to the historical data, facts, context, and argues instead that the text actually precedes that context and blocks access to it.

The historically real, the past real, is that to which I can be referred only by an artifact that is textual in nature. The indexical, iconic and symbolic notations of language, and therefore of texts, obscure the nature of this indirect referentiality and hold out the possibility of (feign) direct referentiality, create the illusion that there is a past out there that is directly reflected in the texts. But even if we grant this, what we see is the reflection, not the thing reflected. (1987: 209) In Metahistory White defines history as "a verbal structure in the

form of a narrative prose discourse that purports to be a model, or



icon, of past structures and processes in the interest of explaining what they were by representing them" (1973: 2).

White provisionally accepts the distinction between history and fiction according to real and imaginary events, objecting only that this obscures the extent to which historians also employ "invention." Fol- lowing Frye, White describes four such explanations by emplotment, Romance, Comedy, Tragedy, and Satire. These are parallel to four explanations by Formal Argument, Formist, Organicist, Mechanistic, and Contextualist, and four explanations by ideological implication, Anarchism, Conservatism, Radicalism, and Liberalism. A historio- graphical style consists of a combination of these levels. White further parallels these groups with the tropes Metaphor, Metonymy, Synec- doche, Irony, which are used by the historian to prefigure the historical field prior to bringing to bear on it the various "explanatory" techniques used to fashion the story out of the chronicle (ibid.: 5). The choice of explanatory techniques is determined by the imperatives of whichever trope informs the linguistic protocol he has used. That is, the tropes of the text to be written precede the historical data to be described. Thus, text precedes context, rather than (in the historians we have examined) context preceding text. White reverses the temporal hierarchy with which he began (ibid.: 7).

In Tropics of Discourse, White develops the theoretical basis of this argument. Not just historical discourse but all discourses in the human sciences reverse this familiar context:text relationship: "We are faced with the ineluctable fact that even in the most chaste, discursive prose, texts intended to represent 'things as they are' without rhetorical adornment or poetic imagery, there is always a failure of intention" (1978: 3). Discourse must therefore be analysed on three levels: the level of description of the data (mimesis); the level of the argument or narrative (diegesis); and the level of the combination of the two. Any discursive analysis of processes of consciousness first projects the four- fold pattern of tropes onto the "data," in order to emplot them, and then charts the growth from naive (or metaphorical) apprehensions of reality to self-reflective (or ironic) comprehension. The result of this analysis is a pragmatic view of historical knowledge: "I have never denied that knowledge of history, culture, and society was possible; I have only denied that a scientific knowledge, of the sort actually attained in the study of physical nature, was possible" (ibid.: 23).

History can achieve another kind of knowledge, "the kind of knowledge which art and literature give us in easily recognizable examples" (ibid.). White has been accused of erasing the boundary between history and fiction, and this he has denied. His argument is that historical contexts are different from fictional contexts, but are not accessible to analysis because historical texts are texts as fictional texts are texts.



Nevertheless, the question may well be asked of White's attempt to write a poetics (or an "historics") of history: What is the point of historical writing, if its traditional function, to represent an historical context, is seen to be self-defeating?

White's argument privileges a particular type of history, intellectual history. Although the text blocks access to the real, a history that took as its object the text itself, i.e., the signifying practices of rhetoric, would be able to employ the self-reflexivity of texts in its favour. Such is intellectual history, and especially White's own semiological historiography. It may well be that historical texts are "reflections, not the thing reflected," but "by directing our attention to the reflection of things that appear in the text, a semiological approach to intellectual history fixes us directly before the process of meaning production that is the special subject of intellectual history conceived as a sub- field of historical inquiry in general" (White 1987: 209). Does White solve the text-context problem by thus making the context immanent to the text? And what of the mass of historical writing that is not concerned with texts but with contexts? One criticism, made by Sande Cohen in Historical Culture, concerns attempts by narrativist historians "to recode story as a cultural universal" (1986: 80). For Cohen, White wrongly assumes that "plot structures can be extracted from a reading of Western literature and correlated with tropes, which makes them repeatable and long-lasting formulas, not transcendental types" (ibid.: 83).

In effect, White's "semiological" history is open to the accusation that has generally been levelled at Structuralism, of extracting the object of its analysis from historical time, as a result of or in order to make it a synchronic entity (i.e., an entity without time, as well as out- side time). In the case of historiography, this is a particularly damning charge. It defeats the whole object of history, which (by any definition, including White's) is to locate its object in history. An argument similar to White's can be made about psychoanalysis, for example-that case histories, which make themselves out to be representations of individual contexts, are in fact the imposition of one particular plot, that of the Oedipus myth, onto those contexts-but psychoanalysis has a number of responses to this accusation, which argue that psychoana- lytic narration is interpretive and performative rather than representational. No such arguments can be produced for historical narrative, so White's argument ultimately condemns all history, including his own, to ahistoric formalism.

The most important point is that rhetoric itself does not exist "out- side" history, and therefore a tropological model cannot be made the basis of a wholly immanent analysis. The codes that White employs -comedy, romance, tragedy, satire-are not universals, but are de-



rived from texts that have their own historical contexts-whether of Frye's criticism or of the specific texts from which Frye derives his plots. This dependence on access to unprocessed historical data is the central characteristic of antinarrativist or objective history, and White does in fact occasionally make comments that only an objective historian would make, for example, "As for revolution, it always misfires" (1987: 227).

Thus, in historiography we are faced with a dilemma: Any analysis that emphasizes context and characterizes its narrative representation as supplementary only invokes the narrative text that blocks the way to the context; this is White's criticism of traditional historiography. But conversely, any analysis that foregrounds narrative without reference to context, or negates or erases that context, as White's semiological history does, only invokes the context that it argued could not be reached; this is Cohen's criticism of White. As David Carroll comments:

To chase away the "ideology of representation" only to replace it with what could be called an idealism of the signifier is in fact to remain within the same metaphysical enclosure. Representation, the text as the expression of a subject's relation to the real, and formalism, the text as an autonomous closed, centred system, are two sides of the same coin. (1982: 18) Carroll's metaphor of the sides of a coin, reminiscent of Saussure's

image of the relation signifier/signified as two sides of the same piece of paper, is slightly misleading. Representation and formalism are two opposed concepts that are closely joined together; but it must be stressed that as two elements in a contradiction, they are oppositions that must be synthesized and yet cannot be synthesized. As Peter Gay puts it, style, which unites form and content, is a centaur, "join- ing together what nature, it would seem, has decreed must be kept apart" (1975: 3). To progress beyond this paradoxical opposition, we need not simply a dialectical synthesis but, rather, what Adorno (1973 [1966]) calls a negative dialectics-in this case, a negative dialectics of historiography. For Adorno, the function of a negative dialectics would be to transcend logical contradictions by way of an antisystem that would neither reduce contradiction to systematic unity or identity nor abandon it to pure difference-that is, in the case of historiography, would treat historical data neither as the object of general laws nor as the uniquely different objects of relativism. For Adorno, contradiction leads neither to identity nor to difference but to non-identity thinking (see Adorno 1973 [1966]: 5); we need therefore to do some non-identity thinking about history.


392 Poetics Today 11:2

II In order to do such thinking, we must look for a historiographer who attempts to go beyond this opposition between representation and reflexivity into post-historicism, who will resolve the problem of the relationship of history and fiction, while at the same time attempting to analyse the nature of historical investigation as a relationship of narrative texts to their contexts of narration. I will argue that Ricoeur proposes such an analysis, but does not, ultimately, achieve it.

Ricoeur's argument differs from the two positions, pro- and anti- narrative, that we have examined.

My thesis, therefore, is equally different from two others-the one that would see in the retreat of historical narrative the negation of any connection between history and narrative, making historical time a construction without any support from narrative time or the time of action; and the one that would establish between history and narrative a relation as direct as that, for example, between a species and a genus, along with a directly read- able continuity between the time of action and historical time. My thesis rests on the assertion of an indirect connection of derivation, by which historical knowledge proceeds from our narrative understanding without losing anything of its scientific ambition. (1984 [1983]: 92) Ricoeur argues that the relationship between time and narrative-

both narrative history and narrative fiction-is of a kind that will not permit the type of hierarchization of text over context, or vice versa, that I have described. This relationship is one that we can call reciprocal presupposition; that is, rather than being in a relationship of supple- mentarity, in which narrative follows the historical time that it is taken to represent, or of anteriority, in which narrative precedes the historical time that it appears to represent, time and narrative presuppose each other. Since the existence of each presupposes the existence of the other, no hierarchy can be established between the two. As Ricoeur puts it: "Time becomes human time to the extent that it is organized after the manner of a narrative; narrative, in turn, is meaningful to the extent that it portrays the features of temporal experience" (ibid.: 3). That is, while time is a non-narrative object, it must be narrated in order to be known; therefore, White's argument that emplotment has priority over historical analysis appears correct. On the other hand, narrative itself must portray temporal experience, that is, must narrate time in order to be narrative; and so the argument that the function of narrative is supplementary to a prior, non-narrative reality is equally correct. This is counterintuitive, to say the least, because these opposed positions cannot both be right, yet it appears that both positions must be right.

Ricoeur's explanation of this paradox is focused on an analysis of mimesis, which he conceives of as a threefold operation: mime-



sis , prefiguration, mimesis2, configuration, and mimesis3, refiguration. Ricoeur's analysis emphasizes configuration, the "plot" as such, conceived as the mediation between the two "sides" of the work, prefiguration, the emplotment of the pretextual reality, and refiguration, the reception of the emploted reality by the reader: "What is at stake ... is the concrete process by which the textual configuration mediates between the prefiguration of the practical field and its refiguration through the reception of the work" (ibid.: 53).

The field of action of mimesis (prefiguration) can only be narrated by mimesis2 (refiguration) because "it is already articulated by signs, rules, and norms." Nevertheless, Ricoeur does talk of a "pre-narrative structure of temporal experience"-hence the reciprocity of mimesis and mimesis2. The continuity of mimesis2 and mimesis3 is attested to by the features of sedimentation and innovation, features which require the support of reading if they are to be reactivated. Again the relationship is one of reciprocity; as in the concept of defamiliarization, the act of innovation presupposes a sedimented or conventional form, while the sedimented form can only be seen as conventional once it has an innovation which deviates from it. To sedimentation must be referred the paradigms that constitute the typology of emplotment. Ricoeur argues that the narrative paradigm itself might be endangered by this process of innovation (ibid.: 70); that is, in a conception of narrative which differs from White's, the narrative paradigm is not ahistorical, but is itself a sedimented structure that could be transformed by innovation.

Mimesis3 is the intersection of the world of the text and the world of the reader, that is, the intersection of the world configured by the poem and the world wherein real action occurs and unfolds its specific temporality (ibid.: 71). The act of reading, as the operator that joins mimesis3 to mimesis2 by way of innovation and sedimentation, is "the final indicator of the refiguring of the world of action under the sign of the plot" (ibid.: 77).

The refiguration of narrative is part of an alternative hermeneutics that does not restore the author's intention behind the text but, rather, analyses the way the text "unfolds" a world in front of itself: "What a reader receives is not just the sense of the work but, through its sense, its reference, that is, the experience it brings to language and in the last analysis, the world and the temporality it unfolds in the face of this experience" (ibid.: 79).

Ricoeur's proposed mediation will be "a new type of dialectic between historical inquiry and narrative competence." Historicism's attempt to subsume individual historical events under general laws ignores the narrative status of the event. No difference is allowed between a historical event and a physical event that simply occurs, and it is not considered relevant that the historical event was recounted


394 Poetics Today 1 :2

in chronicles, or legends, or reports. Therefore, it is not possible to remove history altogether from events and, thus, from narrative. As Ricoeur comments: "If history were to break every connection to our basic competence for following a story and to the cognitive operations con- stitutive of our narrative understanding... it would lose its distinctive place in the chorus of social sciences. It would cease to be historical" (ibid.: 91).

But this does not mean that narrativist explanations are the answer. As we have seen, rejecting the idea of covering-laws in history implies a return to the event, conceived as unique. But this would make any kind of explanation impossible. For Ricoeur, historians need to operate with a particular conception of "uniqueness," one relative to the level of precision of their inquiries (ibid.: 124). Historians are not interested in classifying events into general groups but, rather, in describing how particular events differ from these general groups. His- torical explanation need not be based in either large-scale historical processes or in the actions of individuals (ibid.: 131). The implication here is that while historicism's abstract generalities are not adequate to historical explanation, nor is the narration of individual historical sequences. There must be a mediation attempted between the abstract (systematic) laws of historicism and the contextual (discursive) narration of unique historical sequences. For Ricoeur, emplotment-the "abstract" structures of narrative-may achieve this mediation, but only if the structures of emplotment are dynamic, that is, only if they are not abstracted, but remain linked to discourse by way of temporality. Historiography must employ the ahistorical, achronic structures of narrative that White has described, but it must employ them so as to ensure that they retain both historicity and diachrony and that they remain linked somehow to the historical contexts in which they are generated.

Ricoeur's method is called "questioning back." The analysis involves three aspects: explanation by singular causal imputation, analysis of historical entities, and investigation of the epistemological status of historical time in relation to narrative temporality. These three areas permit mediation between historicism and historical relativism by de- veloping the concepts of quasi-plot, quasi-character, and quasi-time.

Singular causal imputation accomplishes the transition between narrative causality-"the one because of the other"-and explanatory causality, i.e., explanation by laws. Hence, it is "quasi-causal" explanation and involves constructing in the imagination a different course of events than those perceived, then comparing the imaginary consequences with the known consequences. If modifying or omitting an event would, in the historian's judgment, have resulted in a signifi- cantly different course of events, that event can be judged historically



significant. Ricoeur relates this method both to emplotment and to causal laws. It is at once the trying out of different plot schemata and determined by the concept of "objective possibility."

The concept of "first-order entities" is similarly designed to mediate between the entities constructed by the historian (which are abstract and therefore ahistorical) and the characters of narrative fiction. These entities are societal, that is, they are not decomposable into individuals, yet they do refer to individuals-i.e., they are quasi-characters, equivalent to the quasi-plot of singular causal imputation. So, for example, to say "France does this" is not to reduce France to its individuals, yet refers to groups of individuals (Ricoeur 1984 [1983]: 197-99). Similarly, Ricoeur argues, the time constructed by the historian is dependent on the temporality of narrative, and there exists a mediation between them-the quasi-event. The epistemological characteristics of the event-singularity, contingency, deviation-are reformulated in terms of mimesis2. Plots are both singular and nonsingular, contingent and probable, paradigmatic and deviant. The emplotment process oscillates between conformity to the narrative tradition and rebellion with respect to any paradigm received from that tradition. Due to the fact that they are narrated, events, like plots, are singular and typical, contingent and expected, deviant and dependent. Thus the historical event is a quasi-event, neither unique nor universal. This is an example of what Adorno would call non-identity thinking; events as conceived by Ricoeur are neither uniquely different, nor abstractly identical, but non-identical. So, for example, to think of the Russian Revolution as a quasi-event would mean seeing it neither as identical to all revolutions (as a cyclical theory of history would do) nor as uniquely different from all other revolutions (as historicism would do), but would mean interpreting this revolution as the same and not the same as all other revolutions.

Thus the concepts of quasi-plot, -character, and -event establish a mediation between scientific history and individual narration, between historicism and historical relativism, and can do so only by demonstrating that mimesis2-narrative configuration-is not synonymous with fiction. A reciprocal relationship is established between mimesis2 and mimesis that is true both of history and fiction: the representation of mimesis1 is dependent on concepts borrowed from mimesis2, while the function of mimesis 2 is the representation of mimesis1.

Thus, Ricoeur's argument that the structure and context of historical narrative are in a reciprocal rather than a hierarchical relationship allows him to draw a very important conclusion-that there exists a kind of speculative discourse, quasi-narrative, that is neither fiction nor theory. And yet at the very moment when he makes this point, he finds himself forced to reaffirm the hierarchical nature of the rela-



tionship between history and its context. The method Ricoeur is de- veloping, he argues, still presumes the prior existence of the historical "real," which distinguishes this "quasi-narrative" from narrative fiction: "It is for this reason that historians are not simply narrators: they give reasons why they consider a particular factor rather than some other to be the sufficient cause of a given course of events. . . . Northrop Frye is right: poets begin with the form, historians move toward it" (ibid.: 186). No sooner has Ricoeur proposed his radical model of reciprocal interrelationships between the narrative text and its various contexts than he begins to revise it. The projected or unfolded world of the metaphoric text is good enough for narrative fiction, but not for narrative history because the historicity of history is dependent on its past-i.e., mimesis 1, prefiguration, not mimesis3, refiguration. As Ricoeur comments, "Only history can claim a reference inscribed in empirical reality, inasmuch as historical intentionality aims at events that have actually occurred" (ibid.: 82).

This then is the traditional ground for the opposition between history and fiction. The implication here is that history and fiction differ in terms of mimesisl, which fiction lacks; the prefigured field, conceived as the pretextual reality that the text configures, is of more importance to the historical narrative than to the fictional narrative. But this distinction appears to hierarchize the relation of prefiguration and configuration in history writing, where the configuration is once again seen as a supplement to the pretextual reality; and it seems likely that this will in turn hierarchize the relation of configuration and refiguration because in the historical text, unlike the fictional text, the process of refiguration will be supplementary to, and constrained by, the configuration of the pretextual reality. That is to say, whereas fiction lacks mimesis', history will lack mimesis3; the role of the reader in the reception of historical narrative is reduced to nothing. Thus, it seems that the only point of correspondence between history and fiction is mimesis2, emplotment itself.

The reciprocity of time and narrative that Ricoeur had originally intended to describe is not found therefore in the interrelationship of prefiguration and configuration that he actually describes, and he reverses his own radical arguments to return to the familiar hermeneutic model which he has himself criticized.

Having reaffirmed the distinction between history and fiction, Ricoeur turns his attention, in Time and Narrative, vol. 2 (1985 [1984]), specifically to fictional narrative, that is, to the significance of time for and in fiction and the significance of fictional narrative for the understanding of time. As Time and Narrative, vol. 1, has made clear, Ricoeur distinguishes between fiction and configuration: all configurations may be works of the imagination, but Ricoeur reserves the term fiction for



"those literary creations that do not have historical narrative's ambition to constitute a true narrative" (1984 [1983]: 3).

Time and Narrative, vol. 2, contrasts narrative understanding, which results from "an unbroken familiarity with the modes of emplotment throughout history," with the rationality of narrative semiotics (1985 [1984]: 29). By the end of Time and Narrative, vol. 1, Ricoeur has developed the concepts of quasi-plot, quasi-character, and quasi-event to mediate between the abstractions of "scientific" history and the unique stories, people, and events of which history is made up, but about which nothing of an explanatory nature can be said. A similar argument is now made for narrative fiction. Criticism of narrative fiction cannot simply identify itself with literary history, which confines itself to listing individual works in pure contingency. On the other hand, the abstractions of semiotics, which replace the pure difference of literary history with general structures, only do so by removing historical objects from historical time. That is, if we replace the term "event" in the previous discussions of historicity with the term "work," we see that the same paradoxes are true of literary history as well; the work, like the event, cannot be considered unique and yet cannot be subsumed under general laws. Once again, we need to think of literary works in terms of non-identity. The mediation here is "traditionality," which stands midway between a mere history of genres, types, and works, and a logic of narrative.

However, while this proposed mediation corresponds to the one between abstractions and individual events that was called for in Time and Narrative, vol. 1, it is again not achieved. The idea of a discourse, a parole, is privileged over the idea of a system, a langue, and narrative understanding is privileged over narrative semiotics: "The precedence of our narrative understanding, as it will be defended . . . in light of the rationalizing ambitions of narratology, can only be attested to and maintained if we initially give this narrative understanding a scope such that it may be taken as the original which narratology strives to copy" (1985 [1984]: 7). Thus, we immediately have restored the spatio-temporal hierarchy-original and copy-that Time and Narrative, vol. 1, had initially attempted to avoid by positing the reciprocity of mimesis1 and mimesis2.

Ricoeur discusses Northrop Frye's Anatomy of Criticism, arguing that it stems from the transhistorical schematism of the narrative understanding, rather than from the ahistorical rationality of narrative semiotics. Cultures have produced works that may be related to one another in terms of family resemblance, which operate on the level of emplotment; the order among these works is to be assigned to the imagination; as an order of the imaginary, it includes an irreducible temporal dimension, traditionality: "Each of these three points allows



us to see in emplotment the correlate of a genuine narrative understanding that precedes, both in fact and by right, every reconstruction of narrating in terms of a second-order rationality" (ibid.: 19). Ricoeur goes on to consider the position opposite to this one. Does the schematism allow for such variations that the style would no longer be recognisable? The concept of traditionality contains the idea that identity and difference are inextricable. Could the paradigms set up by the self-configuration of the tradition engender variations that threaten identity to the point of causing the death of that style? (ibid.: 19-20).

The reply to this danger-that narrative is in fact a deep-structural model that is fundamentally ahistorical, i.e., that permits historical progression among its variants on the level of individual manifestation while never itself changing-is the touchstone of narrative semiotics. As we have seen, it is against this ahistoricity and in favour of a transhisto- ricity that Ricoeur has developed the idea of narrative understanding. But without the idea of an unchanging, deep-structural model, the possibility of the death of the narrative form is always present. And yet, as Ricoeur comments, "We have no idea of what a culture would be where no one any longer knew how to narrate things" (ibid.: 28).

How can we reconcile the need for the historicity of the narrative understanding with the desire for the continuation of the narrative paradigm? Let us examine Ricoeur's argument against the achrono- logical model of narrative. Considering the classic structuralist arguments of Propp, Bremond, and Greimas, Ricoeur argues that in each case the description of a narrative structure is preceded by narrative understanding. Propp's morphological analysis, like any operation that segments and orders sequentially, is dependent on a prior knowledge of plot as a dynamic unity and emplotment as a structuring operation (ibid.: 38). Similarly, Bremond's analysis of the logic of roles in "elementary sequences" borrows the idea of vectorality from a prior knowledge of plot: "Plot stems from a praxis of narrating, hence from a pragmatics of speaking, not from a grammar of langue" (ibid.: 44). In the same way, Greimas's attempts to do away with diachrony only suc- ceed in underlining the "irreducible role of temporal development," in the concept of the test characterised on the diachronic level as quest. The mediation that the narrative brings about as a quest is not simply logical but really historical. The test, quest, and struggle cannot be reduced to the figurative expression of a logical transformation, for the transformation is the ideal projection of a temporalizing operation.

Whether the story explains the existing order or projects another order, it posits, as a story, a limit to every purely logical reformulation of its narrative structure. It is in this sense that our narrative understanding, our understanding of the plot, precedes any reconstruction of the narrative on the basis of a logical syntax. (Ibid.: 47)



Thus, again, context-the context of narration-is privileged over structure:

The fundamental question raised by the narrative grammar model is whether the so-called 'surface' level is not richer in narrative potential than the deep grammar, and also whether the increasing enrichment of the model as it follows the semiotic traversal does not proceed from our ability to follow a story and our acquired familiarity with a narrative tradition. (Ibid.: 55) The answer to the second question is certainly yes, but I would

argue that the relationship of reciprocal presupposition also exists between narrative understanding and narrative semiotics. Although it is true that the abstractions of narrative semiotics can only exist because they are preceded by a body of individual narratives, the reverse of this, paradoxically, is also true-the existence of a body of individual narratives presupposes an abstract narrative structure which has generated them. Consequently, no hierarchy can be established between narrative semiotics and narrative understanding, as each one presupposes the other. And so the answer to Ricoeur's first question is no: the surface level is not richer in potential than the grammar because they are equally dependent on each other. Ricoeur comes quite close to recognizing this, commenting that "the relationship between the semiotic level and the level of actual praxis is one where each takes precedence over the other. ... In this sense, the surface grammar is a mixed grammar, a semiotic-praxic grammar" (ibid.: 58).

This reciprocal formulation, close to Greimas's concept of a structural semantics, is one with the reciprocal relationship between mimesis and mimesis2 described in Time and Narrative, vol. 1. This reciprocity is further affirmed when Ricoeur goes on to discuss time as a theme of fiction as well as a structurating element of fictional narratives (ibid.: chs. 3 and 4). The particular irony of this argument is that while it invalidates the grounds on which history and fiction have been distinguished by Ricoeur, it simultaneously invokes the need for history, on the grounds that both history and fiction (and narrative in general) must be situated in their context of narration, i.e., their historical context.

How then can we retain the idea of access to an historical context if the traditional argument for the distinction between literature and history does not hold up? How can we account for the reciprocity of the structure and context of narrative history, as Ricoeur has tried and failed to do? And how do we account for the reciprocity of the structure and context of narrative fiction, as Ricoeur has indicated must be done, although he has made no attempt to do so?



III Ricoeur has described a series of oppositions-narrativist and antinarrativist history, representation and formalism, historicism and historical relativism, narrative understanding and narrative poetics-and has revealed these oppositions to be contradictions. He has demonstrated the way in which these contradictory oppositions can be overcome- by recognizing and exploiting the reciprocal presupposition of time and narrative-and then has retreated from this radical formulation because to proceed with this argument would have meant losing the traditional means of distinguishing between history and fiction.

We must find a way to fulfill Ricoeur's original project for a description of the reciprocal interrelations of prefiguration, configuration, and refiguration, which would avoid the problems we have described that arise from the hierarchizing of text over context and vice versa. A result of this reciprocity would be making no distinction between history and fiction on these grounds; fictional narrative would have a prefigured real in the way that historical narrative does (although in neither case would this real be privileged over its configuration), and history would be subject to a process of refiguration in the way that fiction is (again in a relationship of reciprocity).

Although we are no longer so concerned to make a distinction between history and fiction, it is to the theory and practice of the novel that we must turn for a resolution of the problems that have arisen. Here the opposition between representation and formalism can be more clearly seen, and its contradictory nature more clearly revealed. If we can find a way to transcend this opposition in fiction, we can read this argument back onto history.

Henry James gives a classic statement of representation: The only reason for the existence of a novel is that it does attempt to represent life. When it relinquishes this attempt, the same attempt that we see on the canvas of a painter, it will have arrived at a pretty strange pass. [It is necessary] to insist on the fact that as the picture is reality, so the novel is history. (1957 [1888]: 50)

This position is hardly supported by his fiction; in "The Author of 'Beltraffio'," for example, the narrator, who is himself a critic, argues for the opposite position:

That was the way many things struck me in England at the time-as repro- ductions of something that existed primarily in art and literature. It was not the picture, the poem, the fictive page, that seemed to me a copy; these things were the originals, and the life of happy and distinguished people was fashioned in their image. (1986 [1884]: 61)

This comment must be read self-reflexively, despite James's assertion that the story is representational. James argues that "there is surely no



school which ... urges that the novel should be all treatment and no substance" (1957 [1888]: 62). There is such a "school," that of meta- fiction. As Alain Robbe-Grillet argues in contrast to James:

Even if we admit that there is still something "natural" in man's relationship with the world, the fact is that writing, like all the arts, is, on the contrary, ex- traneous to it. The novelist's strength lies in his freedom to invent, without a model. (1965 [1961]: 63) This argument was, of course, the one which historians used to deni-

grate the novel in comparison with the "truth" of historical narrative. In an argument reminiscent of White's characterization of historical writing, Robbe-Grillet asserts the autonomy of the text: "It doesn't lean on any truth that may have existed before it, and one can say that it expresses nothing but itself" (ibid.: 73). But in reversing the hierarchy that James establishes, Robbe-Grillet does not escape "that old leaking ship," the form/content opposition, as he claims to, but merely inverts it: "To talk of the content of a novel, as if it were independent of its form, is to eliminate the whole genre from the domain of art" (ibid.: 71). The content of the novel is to be found in its form; change the order of the words in Madame Bovary, Robbe-Grillet argues, and "nothing is left of Flaubert." This is not, however, to transcend the opposition between form and content but simply to valorize form. When Robbe-Grillet attacks realist criticism for mistakenly attempting to reduce the novel to an extra-textual meaning (ibid.: 71), he replaces James's "copy" model, based as it is in "the poverty of the old myths of 'depth'" (ibid.: 56), with an equally misleading metaphor of "surface." Art is not a "highly coloured envelope whose function it is to contain the author's 'message'," it is the envelope itself (ibid.: 73).

To transcend this opposition we must turn to the Bakhtin circle's theory of "architectonic" criticism. An architectonic is any element of a text that is simultaneously form and content; as the literary text is constituted by such elements, the text is both a representation of reality and an autonomous rhetorical structure which has "refracted" (or configured) any pre-textual real:

The literary structure, like every ideological structure, refracts the generat- ing socio-economic reality, and does so in its own way. But, at the same time, in its "content," literature reflects and refracts the reflections and refrac- tions of other ideological spheres (ethics, epistemology, political doctrines, religion, etc.). That is, in its content, literature reflects the whole of the ideological horizon of which it is itself a part. (Medvedev 1985 [1928]: 17)

As a result, fiction, while not "representing" a pre-textual reality, does nevertheless have a "nonfictional" aspect, which is its historicity- the internalized or refracted ideological horizon. Consequently, if the arguments we have followed concerning the narrativity of historical



writing demonstrate that history is "merely" narrative, the correlate of these arguments is that narrative-even narrative fiction-has an irre- ducibly historical element: "The utterance is not a physical body and not a physical process, but an historical event, albeit an infinitesimal one" (ibid.: 120).

Time, as it is conceived by Ricoeur, is an architectonic element, simultaneously form and content. Time is both a formal element of the text, a structuring principle-as narrative diachrony, it is internal to the text; it dynamizes the apparently achronic structures of narratology-and it is an element of content, a part of the extratextual reality that narratives represent. The reciprocal relationship of time and narrative that Ricoeur originally described is a result of the architectonic nature of time. Time cannot be made the object of either poetics or hermeneutics, nor can it be conceived of as either an extratextual reality or a wholly formalized textual element. It is both of these things together.

Consequently, what is required is not a theory of history that would hierarchize history over fiction but, rather, a dialogism of historical narrative. Bakhtin's own revised theory of genre (1981 [1934]), which opposes the static and achronic conception of genre associated with Formalism to the deconstructive power of the "anti-genre" of the novel, contains such a dialogic theory of history. Bakhtin's attempt to describe the "non-identity" of genre, which is stable and unstable, historical and ahistorical, general and particular, returns historicity to fiction even as it reveals the "novelness" of historical narrative.

References Adorno, T. W.

1973 [1966] Negative Dialectics, translated by E. B. Ashton (London: Routledge). Bakhtin, M. M.

1981 [1934] "Forms of Time and of the Chronotope in the Novel," in The Dialogic Imagination, edited by M. Holquist, translated by M. Holquist and C. Emerson (Austin: University of Texas Press).

Carroll, David 1982 The Subject in Question (Chicago: Chicago University Press).

Cohen, Sande 1986 Historical Culture (Berkeley: University of California Press).

Danto, A. C. 1985 [1968] Narration and Knowledge, revised edition of Analytical Philosophy of

History (New York: Columbia University Press). Davis, Natalie Zemon

1988 Fiction in the Archives (Cambridge: Polity Press). Dray, W. H.

1957 Laws and Explanation in History (Oxford: Oxford University Press). Foucault, Michel

1987 The History of Sexuality: An Introduction, translated by Robert Hurley (Lon- don: Peregrine).



Freud, Sigmund 1962 [1942] Three Essays on the History of Sexuality, translated by James Strachey

(London: Penguin). Gallie, W. B.

1964 Philosophy and the Historical Understanding (London: Chatto). Gay, Peter

1975 Style in History (London: Cape). Hempel, C. G.

1949 "The Function of General Laws in History," in Readings in Philosophical Analysis, edited by H. Feigl and W. Sellars (New York: Appleton-Century- Crofts).

James, Henry 1957 [1888] "The Art of Fiction," in The House of Fiction, edited by Leon Edel

(London: Rupert Hart-Davis). 1986 [1884] "The Author of 'Beltraffio,'" in The Figure in the Carpet and Other

Stories, edited by Frank Kermode (London: Penguin). LaCapra, Dominick

1985 History and Criticism (Ithaca: Cornell University Press). Mandelbaum, Maurice

1977 The Anatomy of Historical Knowledge (Baltimore: Johns Hopkins University Press).

Medvedev, P. M. 1985 [1928] The Formal Method in Literary Scholarship (Cambridge: Harvard Uni-

versity Press). Miller, J. Hillis

1974 "Narrative and History," English Literary History 4(1): 455-73. Mink, Louis 0.

1978 "Narrative Form as a Cognitive Instrument," in The Writing of History, edited by R. H. Canary and H. Kozicki (Madison: University of Wisconsin Press).

Ricoeur, Paul 1984 [1983] Time and Narrative, Vol. 1, translated by K. McLaughlin and D. Pel-

lauer (Chicago: Chicago University Press). 1985 [1984] Time and Narrative, Vol. 2, translated by K. McLaughlin and D. Pel-

lauer (Chicago: Chicago University Press). Robbe-Grillet, Alain

1965 [1961] "Pour un Nouveau Roman," in Snapshots and Towards a New Novel, translated by Barbara Wright (London: Calder).

White, Hayden 1973 Metahistory (Baltimore: Johns Hopkins University Press). 1978 Tropics of Discourse (Baltimore: Johns Hopkins University Press). 1987 The Content of the Form (Baltimore: Johns Hopkins University Press).


Article Contentsp. [383]p. 384p. 385p. 386p. 387p. 388p. 389p. 390p. 391p. 392p. 393p. 394p. 395p. 396p. 397p. 398p. 399p. 400p. 401p. 402p. 403

Issue Table of ContentsPoetics Today, Vol. 11, No. 2, Summer, 1990Front MatterNarratology Revisited: Editors' NoteOn Narrative Studies and Narrative Genres [pp. 271 - 282]Whatever Happened to Narratology? [pp. 283 - 293]Narratology's Centrifugal Force: A Literary Perspective on the Extensions of Narrative Theory [pp. 295 - 307]What Can We Learn from Contextualist Narratology? [pp. 309 - 328]Theories of the Mind: Science or Literature? [pp. 329 - 347]Narrative Tectonics [pp. 349 - 364]Getting Focalization into Focus [pp. 365 - 382]History's Many Cunning Passages: Paul Ricoeur's Time and Narrative [pp. 383 - 403]New BooksA Tale of Two Dictionaries [pp. 405 - 410]The Cognitive Turn in Narratology [pp. 411 - 418]Guides to Narratology [pp. 419 - 427]Narrative Courses [pp. 429 - 435]An Ideological Reading of Narrative [pp. 437 - 442]

New Books at a Glanceuntitled [pp. 443 - 444]untitled [pp. 444 - 445]untitled [p. 445]untitled [pp. 445 - 447]untitled [pp. 447 - 448]untitled [pp. 448 - 449]untitled [pp. 449 - 451]untitled [pp. 451 - 452]untitled [pp. 452 - 453]untitled [pp. 453 - 454]untitled [pp. 454 - 456]untitled [pp. 456 - 458]

Other Books Received [pp. 459 - 460]Back Matter [pp. 461 - 462]

CONNERTY, 1990. Historys Many Cunning Passages. Paul Ricoeurs Time and Narrative

Documents

history of sexuality

jstor terms

role of narrative

narrative authors

narrative techniques

conditions384 poetics

native narrative account

jstor archive