Although the situation varies in different countries or with different perspectives (Rimmon-Kenan 134-35), many recent accounts of the development of narratology tell stories of evolution, either from structuralist narratology to poststructuralist narratology (Currie; Onega and Landa), or from classical narratology to postclassical narratology (Herman “Introduction”), or from structuralist narratology to cultural and historical narratology (Nunning), or from “a strictly formalist poetics” to a “contextualist narratology” (Darby 829), ¹ or from formal investigation to pragmatic, gender-oriented and ideological investigations that go “beyond form” (Fludernik “Histories”), or from traditional narratology to postmodern narrative theory, with the term “narratology” itself seen as obsolete (Currie 6). These stories vary, some even differ in nature, but one idea is shared in common: the decontextualized formal investigation of generic structures has been and should be abandoned, and narratologists should always take into account contexts (of various kinds—see 2.1). But if we examine respectively narratological theorizing and narratological criticism—often occurring since the late 1980s in the same narratological study—a different picture emerges. In terms of narratological criticism, the picture is indeed one of evolution from a text-based investigation subject to formalist limitations to a more valid and fuller investigation that takes into account contexts and readers. In terms of narratological theorizing, however, the picture is more complicated. Postclassical or contextual narratologies have greatly enriched narratological theorizing in various ways (for a most recent good survey, see Fludernik “Histories”), but when the investigation is concerned with generic textual structures and their generic functions, there is usually no room or no need for the consideration of varied specific contexts. Formal narrative poetics (in the shape of newly-established decontextualized structural models), in effect, has appeared continuously in contextual narratologies, which have also drawn quite extensively on classical narrative poetics in contextual criticisms.

In the second edition of Narratology published by the University of Toronto Press in 1997, Mieke Bal writes in the new preface: “Ten years later, the book was still enough in demand to warrant reprinting it. …the demand for the book did make it obvious that it is an instrument functioning in the public domain that I cannot simply take away” (xiii). This essay will explain, mainly in terms of theoretical validity rather than just in terms of public demand, why one “cannot simply take away” formal narrative poetics. It should, however, be made clear that I am not arguing for the validity of the “formalist” or “structuralist” stance. As discussed by hundreds of theorists over the past 30 years or so, the conception of the text as self-contained, detached from context, is indeed invalid, hence now the universal consensus on the necessity to take account of contexts in narrative interpretation. I am only trying to show the continuing importance of formal narrative poetics in approaches to narrative that give great weight to contextual features. The present study, that is to say, is very much a justification of an important aspect of contextual or postclassical narratologies.

This essay primarily aims at revealing that contextual narratologies and formal narrative poetics have nourished each other over the past twenty years or so. It shows that, within and beyond contextualists’ investigations marked by “dual emphasis on poetics and criticism” (Herman, “Introduction” 3), there exists an unacknowledged triple dialogical relationship: (1) the mutually-benefiting relationship between their new formal theorizing and their contextual criticism; in other words, they develop new formal tools that enable new kinds of contextualized interpretations even as those interpretations sharpen those tools; (2) the mutually-benefiting relationship between their new contributions to formal narrative poetics and classical narratology; in other words, their theoretical contributions both depend upon and expand classical narrative poetics; and (3) the mutually-benefiting relationship between classical narrative poetics and contextualized narratological criticism, the former providing technical tools for the latter, which in turn helps the former to gain current relevance.

What Counts as a Contextual/Postclassical Narratology?

      In the early 1960s through 1970s, the term “narratology” had a clear reference: the systematic description of the structures (differentia specifica) of (verbal, fictional) narrative, aimed at establishing a universal grammar of narrative and a poetics of fiction. But since the 1980s, the term “narratology” has been extended to cover the criticism of narratives with narratological tools. As Kathy Mezei observes, “By 1989, feminist narratology had entered another important stage, which saw the transformation of theory and theoretical positioning into praxis” (8). This transformation of theory into praxis is notable not only in feminist narratology, but also, in varying degrees, in other postclassical narratologies. The reason underlying the transformation is not far to seek: since the general academic climate was marked by increasing emphasis on reader and context and since the investigation of generic structures, as will be amply shown below, defies such a consideration, many scholars naturally turned to interpretation or criticism, an area that well accommodates such a consideration. Most postclassical works appear either in the form of narratological analyses of individual narratives or in the form of a combination of “poetics and criticism.”

       While arguing for or agreeing with the broadening of the term “narratology” to cover narratogical criticism, the present study does not subscribe to the tendency to extend the term to 阿narrative studies that engage neither in narratological theorizing nor in narratological practice. A case in point is Sally Robinson’s Gender and Self-Representation in Contemporary Women’s Fiction, which is considered by Mezei as a representative work of feminist narratology (“Introduction” 9-10). Robinson’s study is a radical feminist theory of reading, “concerned with how gender is produced through narrative processes, not prior to them” (198, no.23). Instead of drawing on narratology, Robinson explicitly excludes it through distinguishing her study from those of Robyn Warhol and Susan S. Lanser (ibid.). The distinction between feminist theory/criticism and feminist narratology is bilateral, since, as will be shown below, both Warhol (Gendered) and Lanser (“Towards”, Fictions) have marked off the latter from the former. While both approaches are valuable and indispensable, revealing gender politics from different angles (see Shen “The Future”), it is necessary and helpful to see the difference between them.

      Another case in point is Mark Currie’s “poststructuralist narratology.” Challenging the Kuhnian perspective which sees deconstruction as a linear replacement of structuralism, Currie argues that it would be more realistic to see the new critical approaches since the 1980s as being enabled and resourced by narratology (9-10). Along this line of thinking, deconstruction is treated as a new form of narratology, that is, “poststructuralist narratology”; and the development of narratology becomes one evolving “from deductive science to inductive deconstruction of linguistic knowledge” (46-47). But to view deconstruction itself as a new development of narratology is to overlook the fundamental difference between the two: narratology rests on and operates within narrative conventions, while deconstruction aims at subverting narrative conventions altogether.

      In terms of philosophical positions, it is commonly held that Saussure’s emphasis on the relational nature of language in Course in General Linguistics lent much force to Derrida’s theory of deconstruction. But in effect, in Course in General Linguistics, there are two superficially contending forces at work. One pays great attention to the relation between the signifier and the signified, defining language as “a system of signs in which the only essential thing is the union of meanings and sound-images, and in which both parts of the sign are psychological” (Saussure 15, my empahsis). The other force views language only as a system of “differences. Even more important: a difference generally implies positive terms between which the difference is set up; but in language there are only differences without positive terms” (Saussure 120). Indeed, a Western language consists of signs that are in general totally arbitrary, hence by no means positive terms. But we have to be aware that differences alone cannot generate signification. In English, “sun” (/sΛn/) can function as a sign not only because of its difference from other signs in sound or “sound-image,” but also because of the conventional union between the sound-image “sun” and the signified concept. Given, for instance, the following sound-images “lun”(/lΛn/), “sul” (/sΛl/) and “qun” (/kwΛn/), although each can be identified by its difference from the others, none of them can function as a sign, because there is no established conventional “union between meanings and sound-images.” When commenting on Saussure’s theory of language in Positions and other works, Jacques Derrida pays attention exclusively to Saussure’s emphasis on language as a system of differences among the signifiers, to the neglect of Saussure’s emphasis on the relation between the signifier and the signified. As we all know, Saussure in Course in General Linguistics distinguishes three arbitrary relations in the formation of language: (1) the arbitrary system of differences among signifiers; (2) the arbitrary system of differences among signifieds (the way that languages cut up meaning into individual signifieds is arbitrary and varies from language to language), and (3) the conventional connection of a given signifier to a given signified. Since Derrida does not pay attention to (3), the connection between (1) and (2) cannot be established for the simple reason that (3) functions as the only and the indispensable link between (1) and (2). Without (3), language becomes a play of signifiers themselves, which cannot be connected to any signifieds, and meaning naturally becomes forever indeterminable. Now, it goes beyond the scope of the present study to investigate further the philosophical positions of structuralist narratology and deconstruction. Suffice it to say that the two hold opposite views towards the signifying function of language and narrative conventions, and should not be treated as two successive stages of development.

      This, however, does not mean that narratologists are not influenced by poststructuralism. Under the positive influence of poststructruralism, narratologists in general have become more realistic and have modified the original tone of objectivity, certainty, or finality. They acknowledge that any narratological model is only a “heuristic tool” (Bal, Narratology xiii), “a semiotic system of interdependent concepts and terms” (Kafalenos 41), or “a conceptual framework, a set of hypotheses having explanatory power” (Rimmon-Kenan, Narrative 140). And under the positive influence of deconstruction, some narratologists have directed attention to various kinds of uncertainty or undecidability in literature. Rimmon-Kenan, for instance, investigates “undecidability concerning the narrator’s identity and structural position vis-a-vis the events narrated” (A Glance 3-4) in a type of twentieth-century experimental fiction. But as the term “structural” indicates, Rimmon-Kenan acknowledges the existence of the “manipulation of narrative levels” or “specific strategies of storytelling” (3), which forms a striking contrast to Staney Fish’s poststructuralist view that “formal patterns are themselves the products of interpretation” (Is There 267; for a response, see Shen “Stylistics”). 

       In terms of Mark Currie’s account, a notable consequence of the story of evolution from structuralist narratology to poststructuralist narratology is to exclude the real “narratology” from recent scene: “In short, poststructuralists moved away from the treatment of narratives (and the language system in general) as buildings, as solid objects in the world, towards the view that narratives were narratological inventions construable in an almost infinite number of ways” (Currie 3, my emphasis). What we have here is in effect a story of the subversion or replacement of narratology by poststructuralism. Narratology, however, has survived and has been developing together with poststructuralism. The survival of narratology is made possible primarily by the coming into existence of various contextual or postclassical narratologies. The point is that a contextual or postclassical narrative study should not be considered “narratological” unless it asks “narratological questions” or uses “narratological methods and analytic categories” (Rimmon-Kenan, Narrative 144). Only in this way, can we perceive how contextual or postclassical narratologies and formal narrative poetics have been benefiting from each other.

Feminist Narratology

       Feminist narratology has played a pioneering and significant role in preserving and enriching formal narrative poetics. Feminist narratology came into being in North America in the 1980s when structuralist narratology was very much excluded from the scene by the joint forces of poststructuralist and sociohistorical/political approaches. Under such circumstances, both Susan S. Lanser and Robyn Warhol, the two founding and leading figures of feminist narratology, helped to save formal narrative poetics through (1) arguing for its usefulness and relevance, (2) enriching formal narrative poetics, (3) applying classical narratological tools in the criticism of narratives in sociohistorical contexts.

(a) Theoretical Defense

      Faced with a situation where “no contemporary theory, whether Anglo-American or continental, has exerted so little influence on feminist criticism or been so summarily dismissed as formalist-structuralist narratology,” Lanser explores “the compatibility of feminism and narratology” (“Toward” 611) and argues that the “necessarily semiotic nature of even a revised narratology will help to balance feminist criticism’s necessarily mimetic commitments. The comprehensiveness and care with which narratology makes distinctions can provide invaluable methods for textual analysis” (ibid. 614).

      Another defense of narratology comes from Robyn Warhol who, in her Gendered Interventions, asserts that narratology “can do what feminist aesthetic criticism, for example, cannot do: describe exactly what the conventions of fictional discourse are and how they operate” (13). That is to say, “narratolgy” can help describe “differences that might occur among the structures in men’s and women’s texts,” which “would be the first step in developing a poetics of gendered discourse” (15).

Significantly, the defenses from Lanser and Warhol display quite consistent omission of the epithet “structuralist.” These are defenses of narratology as formal investigation of generic textual structures (formal narrative poetics), not of the “structuralist” philosophical stance. And this is a position the present study shares.

(b) Theoretical Preservation and Enrichment

      While defending (structuralist) narratology, feminist narratology, as a contextualist approach, is unequivocally critical of purely formal narratological investigations. The criticism centers on two related issues: (1) being gender-blind, and (2) decontextualization. In The Narrative Act, a study attempting “to forge a feminist poetics of point of view”(Lanser, “Towards” 611), Lanser criticizes structuralist narratology for, among other things, an “adherence to a supposedly value-free methodology; and most critically, an isolation of texts from extraliterary contexts and from their ideological base” (39). Holding a similar contextualist stance, Warhol says,

[Genette] never hints at the possibility of any gender-based differences or patterns among narrative structures. Neither Gerald Prince nor Mieke Bal, in their less specific and more comprehensive presentations of narratology, mentions gender as a factor influencing the models they describe. The oversight is not a sexist one: not only gender, but all variables of context remain outside of classical narratology’s realm. As proponents of structuralism, the first practitioners of narratology lifted texts out of their contexts in order to distill from them the essential structures that characterize all narrative” (Gendered 4).

Warhol appreciates Mieke Bal’s shifting “the emphasis of her scholarship” from decontextualized classical narratology “to semiotics, where she can study recurring textual signs and structures in the context of the cultures that produce them” (ibid.).

      But a close look at the theoretical distinctions Lanser and Warhol themselves propose will reveal that the investigation of generic structures, in striking contrast with narratological criticism, defies, by nature, the consideration of specific sociohistorical contexts. Attention will be directed first to two distinctions made by Lanser: (1) that between public and private narration, and (2) that between authorial, communal and personal voice. Concerning the former, Lanser writes, “By public narration I mean simply narration (implicitly or explicitly) addressed to a narratee who is external (that is, heterodiegetic) to the textual world and who can be equated with a public readership; private narration, in contrast, is addressed to an explicitly designated narratee who exists only within the textual world” (“Towards” 620). Not surprisingly, Lanser’s distinction is very much decontextualized and gender-indifferent. Indeed, as far as such structural classifications themselves are concerned, diversified specific contexts only form irrelevant digressions. As Lanser points out, “a major benefit of narratology is that it offers a relatively independent (pre-textual) framework for studying groups of texts” (“Towards” 611). While the investigation of groups of texts (as communicative acts) needs to take account of the varied contexts in which the texts are produced and interpreted, the establishment of the relatively independent or pre-textual framework necessarily requires lifting texts (as structural illustrations) “out of their contexts in order to distill from them the essential structures” concerned. As for the distinction between the different kinds of “voice” (Lanser, Fictions 16-21), while there is only one “personal voice” and one “authorial voice” in the abstract form with “conventionally distinct modes of authority” (Lanser, The Narrative 137), the reasons underlying the choice of a mode in different narratives may vary from context to context. Lanser’s penetrating and illuminating investigations well reveal how the interaction between decontextualized structural properties and contextualized gender differences determines the choice of a narrative mode in sociohistorical contexts.

Significantly, Lanser’s narratological theorizing helps not only to preserve but also to enrich formal narrative poetics. A most valuable enrichment is found in Lanser’s distinction of “communal voice,” which is “a category of underdeveloped possibilities that has not even been named in contemporary narratology” (Lanser, Fictions 21). By “communal voice,” Lanser means “a practice in which narrative authority is invested in a definable community and textually inscribed either through multiple, mutually authorizing voices or through the voice of a single individual who is manifestly authorized by a community” (ibid.) Lanser distinguishes three forms of “communal voice”: “a singular form in which one narrator speaks for a collective, a simultaneous form in which a plural ‘we’ narrates, and a sequential form in which individual members of a group narrate in turn” (ibid.). The “communal voice” was previously neglected probably because, “[u]nlike authorial and personal voices, the communal mode seems to be primarily a phenomenon of marginal or suppressed communities” (Lanser, Fictions 19). Lanser discovered this mode through her investigation of women’s texts. While the use of this mode by women writers undoubtedly calls for contextualized analysis, the theoretical classification of this mode and its different forms (singular, simultaneous and sequential) requires lifting texts out of their varied specific contexts “in order to distill from them the essential” structural properties concerned. Thus, although the authorial mode has been used more frequently by men and the communal mode mostly, if not only, by women in sociohistorical contexts, Lanser’s theoretical distinction in itself does not reflect this gendered or culture-determined difference. It is true that Lanser does mention “narrative authority,” but what is involved in such abstract theoretical distinctions is only “conventionally distinct modes of authority.” The gender-indifferent and decontextualized structural distinction leaves room for the investigation of “communal voice” in men’s texts – at least it may be found in texts by colored and/or working-lass men. And even if men writers—of whatever class or color—in the past have not yet used this mode, certain men writers in the future may do so for various reasons. It should be stressed that, as in the case of “authorial” or “personal” voice, we recognize “communal” voice or its “singular,” “simultaneous” and “sequential” forms not because of the gender of the author, nor because of any given sociohistorical context, but because of the “essential” structural properties involved.  

      The discovery by Lanser of this mode points to the importance of studying women’s texts in order to build up a more comprehensive narrative poetics. Given the fact that this mode has appeared more frequently, if not exclusively, in women’s texts, not paying sufficient attention to women’s texts can easily lead to the neglect of this mode. Although most narrative structures and techniques are shared by men’s and women’s texts, women writers in sociohistorical contexts may use much more frequently certain techniques and may produce certain types of plot that are not shared by men’s texts (Lanser “Towards”; Page). Thus the investigation of women’s narratives may lead to the discovery not only of neglected modes of narration, but also of new types of plot.

It should, however, be stressed that, insofar as the classification of generic structure is concerned, individual narratives merely serve as formal illustrations. In making the distinction between plot of resolution and plot of revelation, Seymour Chatman has chosen only two women’s texts for illustration: Jane Austen’s Pride and Prejudice for plot of resolution, and Virginia Woolf’s Mrs. Dalloway for plot of revelation (Story 47-8). This choice of women’s texts, however, does not make Chatman’s classification any more feminine than previous models. The theoretical distinction is purely formal, and the two women’s texts merely function as illustrations of “essential” structural properties, serving the same illustrative role as men’s texts of the same types. In order to build up a comprehensive narrative poetics, one should not neglect any type of texts – whether by men or women, white or colored, upper-class or working-class writers. On a wider scale, “general narratology” has been, and can be, continuously enriched by extending its scope of investigation into new fields and new media (see Rimmon-Kenan, Narrative 137-47; Sternberg 300; Chatman Coming; Prince, “On Narratology” 79, Phelan & Rabinowitz).

      Attention will now be directed at Robyn Warhol’s distinction between distancing and engaging narrators. In Gendered Interventions, Warhol writes:

A narrator who provides so much information about the narratee that the addresee becomes, as Prince says, “as clearly defined as any character” necessarily places a distance between the actual reader and the inscribed “you” in the text (“Introduction” 18). Such a narrator I call distancing. But not every narrator who intervenes to address a narratee does so to set the actual reader apart from the “you” in the text. Another kind, which I call engaging, strives to close the gaps between the narratee, the addressee, and the receiver... (29, my emphasis)

Again, not surprisingly, this distinction is as agendered and decontextualized as classical narratological distinctions. Warhol makes it clear that she “would not claim that engaging strategies are specific to women’s texts, nor that distancing strategies occur exclusively in men’s” (17). While the theoretical classification is agendered and decontextualized, the actual choice and usage of those narrative strategies are related to gender differences and determined by sociohistorical contexts. Warhol finds that the “distancing” strategies dominate novels signed by men and the “engaging” strategies dominate novels signed by women in the mid-nineteenth century (17), a preference related to the different ends the women and men used the discourse of realist fiction in the socio-historical contexts (18).

      In Warhol’s study, there can be clearly seen not only the borderline between decontextualized “terminological distinctions” and contextualized criticism, but also the demarcation between decontextualized generic rhetorical effects and the specific contexts of interpretation. After showing how the engaging strategies in Stowe’s Uncle Tom’s Cabin failed to engage a critic, Warhol observes:

Discrepancies such as this between narrators’ moves and audiences’ responses warrant mention here: strategies are rhetorical features of texts, choices of technique indicating novelists’ apparent hopes about the emotional power their stories might wield. Strategies can misfire; they guarantee nothing. A reader’s response cannot be enforced, predicted, or even proven. …The participial forms of the terms distancing and engaging are not meant to imply an action that a text or a narrator could take upon a reader, but rather to identify the rhetorical moves these strategies represent. (Gendered 25-26, original italics and my boldface)

While feminist critics tend to lay emphasis on feminist reading positions (Fetterley, Schweickart, Robinson), Robyn Warhol, as a feminist narratologist, sets store by the narrative strategies and their generic rhetorical effects. This emphasis is shared by Lanser, who is concerned with what effects a narrative technique “can generate,” with its “potential for reader manipulation” (The Narrative 28-9). The reader’s recognition of the generic function of a narrative technique very much depends upon decontextualized literary or narrative competence (Culler 113-30; Prince, Dictionary 65). If an actual reader does not grasp the generic effect of a narrative technique, it will be taken as a misunderstanding of the rhetorical move by the novelist; and in terms of the novelist, it will be taken as a failure of her application of that technique.

      However, the narrative conventions upon which the generic effects of narrative techniques partly rest may not be shared by all times. As we know, narrative techniques are produced and used in sociohistorical contexts. For instance, the frequent appearance of variable internal focalization instead of “omniscience” in modern fiction has to do with the more skeptical and more individualized sociohistorical context after the First World War. If such a modern technique were used in a literary work set in the Middle Ages, it would most probably be out of place, since the narrative conventions (connected with people’s view of the world) were quite different back then (cf. Chatman, Coming 198-99). Nevertheless, once a narrative technique comes into existence, its generic effects and the relevant narrative conventions usually remain quite stable over a long period of time, since both are associated with the “essential” structural properties of the technique. Not surprisingly, the “distancing” and “engaging” effects associated with Warhol’s dual distinction or the different degrees of authority associated with Lanser’s tripartite distinction³ have remained somewhat unchanged and may remain fairly stable in the foreseeable future.

(c) The Case of Sex as a Formal Category

Interestingly, Susan Lanser’s study of women’s texts has led her to formalize and decontextualize “sex” as a narratological category. In “Sexing the Narrative,” Lanser writes,

Written on the body leads me to recognize that sex is a common if not constant element of narrative so long as we include its absence as a narratological variable. Such an inclusion allows us to make some very simple formal observations about any narrative: that the sex of its narrator is or is not marked and, if marked, is marked male or female, or shifts between the two….While the narrator’s sex is normally unmarked in hetetodiegetic texts, sex is an explicit element of most homodiegetic, and virtually all autodiegetic, narratives of length….One might well classify heterodiegetic and homodiegetic narratives according to their marking or non-marking of sex and according to the ways in which sex gets marked: overtly, through explicit designation, or covertly, through conventional aspects of gender that suggest but do not prove sex (87, original italics and my boldface).

This theoretical distinction of “sex” is, not surprisingly, as formal and decontextualized as classical structural distinctions. In the case that the narrator’s sex is unmarked or marked only covertly, the readers’ inference of the narrator’s sex may vary from individual to individual or from context to context, but the theoretical distinction between “marked” and “unmarked” or between “covertness” and “overtness” has to be made in an abstract and decontextualized way. This case lends strong support to the point I have been driving at: the classification of generic structures defies, by nature, contextualization or requires, by nature, leaving aside the consideration of sociohistorical contexts. In a similar fashion, we could formalize the narrator’s race, class, religion, ethnicity, education, marital status or sexual preference, all of which can be either “marked” or “unmarked,” and if marked, can be marked either “overtly” or “covertly” in the text. Once an attempt is made to theorize those non-structural elements as “narratological” categories, it also becomes necessary to lift the texts out of their contexts and to distill from them the distinguishing properties concerned. Such non-structural elements, that is to say, cannot enter the realm of narratogical theory unless they are transformed into decontextualized formal distinctions. Now, contextualist narratologists “argue for the need to inquire into the intentions, motivations, interests, and social circumstances of real authors and audiences. Failure to make this kind of inquiry, they believe, dooms narratology to a treatment of narrative as a ‘detached and decontextualized entity’” (Chatman, “What” 314). But as far as the investigation of generic structures is concerned, there is, in effect, nothing wrong with decontextualization, since this is the only possible way to do it—even in the case of “sex.”

(d) Transforming Narratological Criticism

       Because of formalist limitations, earlier structuralists severed texts from their contexts of production and interpretation in narratological criticism. This defect was fortunately redressed by contextualist narratologies, with feminist narratology taking the lead. Without this timely transformation, narratological criticism could not have survived at least in North America.⁴ By successfully combining a narratological perspective with historical and ideological issues, contextualist narratologies “counter the view that narratology’s formalism entails its futility in the face of social concerns” (Bal, “The Point” 750). As demonstrated by numerous brilliant feminist-narratological analyses (see, for instance, Bal Death; Warhol Gendered and “The Look”; Lanser Fictions; Mezei Ambiguous; Case Plotting and “Gender”), the “more precise such an analysis is, the better it helps to position the object within history” (Bal, “The Point” 750).

      Feminist narratologists, while proposing new structural distinctions such as those mentioned above (see also Warhol “Neonarrative”), have been drawing extensively on classical narrative poetics. If 1989 “saw the transformation of theory and theoretical positioning into praxis” in feminist narratology, the narratological tools used in the praxis are for the most part classical ones. Such applications on the one hand confirm the validity and usefulness of classical distinctions and on the other help classical models to gain current relevance. In September 2002, Routledge published the second edition of Rimmon-Kenan’s Narrative Fiction: Contemporary Poetics, after having reprinted it eight times since its first appearance in 1983; and 2005 sees the publication of The Routledge Encyclopedia of Narrative Theory, which provides “ample coverage of structuralist models” (Herman et al. x), implying the usefulness and current relevance of formal narrative poetics. A significant contribution contextualist narratologies have made to formal narrative poetics is that they demonstrate how decontextualized and gender-indifferent pre-textual frameworks can prove highly useful in contextualised criticism.

      In short, feminist narratologists borrow from and make new contributions to formal narrative poetics. Their investigations, like contextual investigations in general, actually develop a triple dialogical relationship between formal narrative poetics and contextualized criticism as specified in the last paragraph of the beginning part of this essay.

Cognitive Narratology

      In the narratoloical field, as in many other fields such as linguistics and stylistics, the past ten years or so have witnessed a notable cognitive turn, giving rise to a thriving contextual approach “cognitive narratology.” As distinct from feminist narratology that sets store by real authors in sociohistorical contexts, the contextualization of cognitive narratology is oriented towards readers’ mental models in the reception process (see, for instance, Herman Narrative). To clarify the nature and different strands of cognitive narratology, attention will first be directed to a distinction between different kinds of context and different kinds of audience.

(a) Predominance of “Generic Context” and “Generic Audience”

      Basically, contexts can be divided into two kinds, one is generic or conventional, and the other specific or sociohistorical. In terms of the former, we may take a brief look at speech-act theory. Speech-act theory is often concerned with the forces of verbal acts in conventional situational contexts, such as a classroom, a church, a court-trial, a conversation, trying to find a gas station, a newspaper article, a novel, an avant-garde novel etc., where the speaker and the addressee are stereotyped social roles (a teacher, a student, a driver, a priest, a judge etc.). Such contexts are, or are implicitly treated as, agendered and ahistorical. As illustrated by Mary Louise Pratt’s Towards a Speech Act Theory of Literary Discourse, when discussing the generic structures, functions and conventions of (types of) literary discourse, the exclusive concern with the generic situational contexts (e.g. the literary speech situation) is right and proper. But when it comes to the interpretation of a given work, the specific sociohistorical contexts of production and reception need to be taken into account.

      Insofar as narrative is concerned, the two different kinds of context call into being two different kinds of audience: one is what I would like to call “generic audience,” that is, audience who are equipped with “narrative competence” and who share the same narrative conventions as typically embodied by stereotypic assumptions, expectations, frames, scripts, plans, schemata, or mental models in narrative comprehension. The other can be termed “the audience of a given narrative,” comprising four reading positions as first classified by Peter J. Rabinowitz in “Truth in Fiction”: the flesh-and-blood audience (the reading position related to a reader’s particularity and social identity), the authorial audience (the postulated or implied reading position, which is aware of the fictitiousness of the work), the narrative audience (that part of the reading consciousness which treats the fictional world as real), and the ideal narrative audience (the reading position “for which the narrator wishes he were writing”) (134).⁵ Apparently, what I call “generic audience” excludes the consideration of flesh-and-blood individuals across contexts, and is to be distinguished from the authorial and narrative audience of a given narrative in emphasis or focus.

       To gain a full picture, we may go further to distinguish the following three types of study of interpretations:

1)    The study of the understanding of (a type of) narrative as a genre, or of certain structural features as generic features.

2)    The study of the interpretation of the thematic significance of a given narrative (or the function and effect of textual features in relation to the thematic significance of a given narrative). It needs to consider the contexts of production and interpretation, including the different responses of flesh-and-blood individuals.

3)    The study of real people’s understanding of real events as “narrative” in the real world. Depending on whether the study is to reveal general cognitive characteristics or different individuals’ different perspectives, the study will focus either on shared conventions (an individual as a representative of (a category of) people in general) or on different individuals’ particularity.

      It is significant to note that cognitive narratology in general focuses on the first kind of interpretation, which involves only the “generic audience” and the “generic context” of narrative reception, leaving aside the varied sociohistorical contexts. Even if the investigation is based on empirical experiments, cognitive narratologists tend to try to uncover from varied reception processes shared models or mechanisms of narrative comprehension (see, for instance, Ryan, “Cognitive”).

(b) Preserving and Enriching Formal Narrative Poetics

       Although cognitive narratological investigations vary in theoretical heritage, modus operandi, scope, focus, exemplification or the analytical tools adopted, they in general have the following feature in common. While investigating the functioning of interpretive strategies in narrative understanding, the analysts acknowledge and pay great attention to the role of textual structures, since in their view narrative comprehension is fundamentally a process of (re)constructing storyworlds on the basis of textual cues and the inferences that the cues make possible (Herman, Story 6).⁶ Cognitive narratology, that is to say, may be characterized as “the investigation of mental processes and representations corresponding to the textual features and structures of narrative” (Bortolussi & Dixon 24). Precisely because of the importance of “textual features and structures,” formal narrative poetics helping gaining a better understanding of this aspect is complementary to the cognitive approach focusing on audience’s reception. The latter approach in its turn complements the former by revealing various kinds of “interesting cognitive mechanisms” (Jahn 168). Despite the significant cognitive advancement, one narratological aspect has remained unchanged: the models built up by cognitive narratologists to account for “the textual features and structures” are usually as formal and decontextualized as classical structural models, thereby constituting, in effect, a preservation and development of formal narrative poetics.

        I’ll start with David Herman’s Story Logic, which rethinks “narrative as a strategy for creating mental representations of the world” (5), and which, while emphasizing the contextualization of formal description, well demonstrates the complementary relation between decontextualized narratological theorizing and contextualized narratological criticism. The complementary relation is particularly notable in chapter 9 entitled “Contextual Anchoring,” a concept defined as “the process by which cues in narrative discourse trigger recipients to establish a more or less direct or oblique relationship between the stories they are interpreting and the contexts in which they are interpreting them” (8). Using Edna O’Brien’s 1970 novel A Pagan Place as a case study in second-person narration, the chapter focuses on the second person “you” as a special case of person deixis. Herman observes, “In some cases, at least, narrative you does not simply or even mainly refer to storyworld participants but also (or chiefly) addresses the interpreter of the narrative. And sometimes a single instance of narrative you both refers and addresses. The result then is a fitful and self-conscious anchoring of the text in its contexts” (332). While the actual occurrences of narrative you in specific narratives are anchored in contexts, once an effort is made to classify the different types of you as a theoretical framework, the cognitive narratologist is left with no choice but to lift the texts out of their contexts in order to distill from them the structural properties concerned. Herman offers a comprehensive classification of five types of you in second-person fictions: (a) generalized you, (b) fictional reference, (c) fictionalized (=horizontal) address, (d) apostropic (=vertical) address, (e) double deictic you (345). The different types are all determined solely by “essential” structural properties. The first two, for instance, are “marked by an uncoupling of the grammatical form of you from its deictic functions” (340). Type (a) is impersonal, generalized, “pseudo-deictic,” or “non-deictic” (ibid.). Type (b), in contrast, refers to the narrator-protagonist by what Margolin calls “deictic transfer”. In making the classification, the various examples Herman cites from A Pagan Place only serve as structural illustrations, playing the same illustrative role as the hypothetical example offered on page 341 (“When you’re hot, you’re hot—”).

      In Herman’s classification, each type of you has its generic function, which is shared by different occurrences of the same type across contexts and which is therefore to be distinguished from contextualized significance. Further, leaving aside the generalized you, a type that can be found in all kinds of narration, the generic functions of the other types of you are based, to a certain extent, on the generic functions of (certain types of) second-person narration. Precisely because Herman’s classification of the different types of “you” is based on generic textual features without being affected by the diversity of readers and contexts in the same genre, the classification yields valuable “new tools for the poetics of second-person fiction” (Herman, Story 337).

       In Herman’s study, as in cognitive narratological investigations in general, what is focused on is the first type of interpretation as classified above: the understanding of (a type of) narrative as a genre, or of certain structural features as generic features. As far as this type is concerned, a reception-oriented approach has the capacity for seeing the interaction among textual cues, generic conventions and generic interpretive strategies (mental models, frames, scripts, schemata etc.), which are very much interrelated or mutually dependent. Textual cues are a result of the author’s writing the text in a certain way based on generic conventions and with the relevant reading strategies in view. Generic conventions are given rise to by a joint function of generic textual features (as produced by authors whose writing helps make a new genre or expands an existing genre) and generic interpretive strategies (adopted by readers and expected by authors). And generic interpretive strategies are based on generic textual features and generic conventions.

      Primarily because of the capacity for taking account of the interaction, cognitive narratology has been in recent years widely hailed as an approach superior to the text-based formal narrative poetics. In Pcychonarratology, Marisa Bortolussi and Peter Dixon offer a comprehensive summary of various reception-oriented approaches, concluding, “In all these disciplines, this emphasis on the recipient of narratives can be seen as the result of a paradigm shift that exposed and transcended the limitations of purely formalist models” (2, my emphasis). Moreover, Bortolussi and Dixon take issue with other reader-oriented scholars for failing to carry out an empirical approach: “In general, to expand on what might plausibly be attributed to the reader, narratologoists and reader-response theorists have generated a hypothetical description of readers’ knowledge and inferences with little grounding in objective evidence….The result is a circular kind of logic: The characteristics of a text provide evidence for various narrative competencies, and the existence of a particular competence provides the evidence for a particular characteristic of the text” (168). What they advocate is to study “actual, real readers, and to ground one’s analysis of the reading process in empirical evidence on how readers process narrative forms” (168-69). Interestingly but not surprisingly, while theoretically excluding the text-based approach and the “generic reader”-oriented approach, both are adopted in Bortolussi and Dixon’s own three-step investigation: “we first provide a framework for understanding the relevant textual features; then we discuss some hypotheses for related reader constructions; and finally, we report some empirical evidence that supports these hypotheses. Each of these aspects is taken up in turn” (184-185).

      While formal narrative poetics is only concerned with generic textual structures (and their generic functions) with no capacity for dealing with actual readers’ responses in varied contexts, an empirical approach to varied actual readers’ responses, similarly, will be hard put to it to work out models of generic textual structures (but of course, one can uncover shared mechanisms of narrative reception). The failure to see clearly this division of labor is a fundamental reason underlying many criticisms of formal narrative poetics. While criticizing Gérard Genette for failing to take into account “the type of reader, the nature of the text, and pragmatics of the reading context” in his theoretical discussion of focalization (Bortolussi and Dixon 177-78), Bortolussi and Dixon’s own theoretical discussion of focalization, not surprisingly, is just as reader-free and context-free. Their “psychonarratological approach” has synthesized, “from the relevant scholarship in narratology and linguistics,” three categories as the theoretical framework of focalization: (1) descriptive reference frames, (2) positional constraints, and (3) perceptual attributions (186-189). The first category is further divided into “relative reference frames” (e.g. “Some times a dog would howl in the distance”, where we have “perceptual information relative to the location of a potential perceiver”) and “external reference frames” (e.g. “The lamps shone from the summits of their tall poles,” where the reference frame is “determined by axes found in the story world, independent of any potential perceiver”). The second category “positional constraint” is the textual “constraint on the location of an agent who might have perceived the information,” while the third category “perceptual attribution” consists of textual cues “suggesting a perceiver.” Whatever the modification of earlier models, Bortolussi and Dixon’s classification of textual features is as decontextualized as classical narrative poetics. What is crucial is to preserve the borderline between “the objective features of the text” and “the potentially variable reader constructions” (Bortolussi and Dixon 198), a borderline that is kept quite clear in Bortolussi and Dixon’s own three-step investigation: “Having discussed some categories of textual features related to perceptual information, the important question that should be considered is how such cues are processed by readers. …We present here several ideas concerning one aspect of the representations readers may construct in processing perceptually salient descriptions. Subsequently, we report some evidence in support of these ideas” (191, my emphasis).

       The three-step investigation well demonstrates the co-validity of, and the mutually-benefiting relationship among, the different kinds of inquiry: (1) the decontextualized investigation of generic textual structures; (2) the investigation of hypothetical generic reader’s understanding of narrative, and (3) the empirical study of actual readers’ cognitive processes. The first decontextualized approach, which provides “a stable landing” or “a theoretical bedrock” (Phelan and Rabinowitz 1), paves the way for the latter two. The third, that is, the empirical, verifies or challenges the conclusions drawn by the second concerned with hypothetical narrative understanding. And the hypothetical and the empirical can shed light on the limitations of the first, promoting its further development.

“Strong” Versus “Weak” Contextualist Position

       In order to gain a better understanding of the complementary relation between contextualist narratologies and formal narrative poetics, we need to examine further the contextualist position. The contextualist position can be divided into the strong version and the weak version. The contextualist investigations as discussed in the preceding part of the essay all fall into the “weak” category. Significantly, the difference between the “weak” and the “strong” versions is not a matter of degree, but of kind. The strong-contextualist position grants the context all the determining power, while the weak version acknowledges the generic identity of the text itself. A vigorous recent attempt to assert the strong version is found in Michael Kearns’ Rhetorical Narratology, which takes speech-act theory as its theoretical ground. Kearns’ strong contextualist claims are one extreme of the contextualist side. But Kearns’ book is valuable and revealing precisely because Kearns’ practice does not fully conform to his theoretical claims. Through analyzing Kearns’ argument in detail, we can see the unavoidable conflict between the strong and the weak position, a conflict that the introduction of speech-act theory paradoxically helps reveal.

      According to Kearns, “the right context can cause almost any text to be taken as narrative and … there are no textual elements that guarantee such a reception” (2).⁷ He cites “an excellent example” from Petrey: “[t]he constitution is suspended” in a newspaper article compared to the same sentence in a government decree (Petrey 12) to illustrate the point that the same string of words can have entirely different effects across different contexts (Kearns 11). In one context, the words are performative, in the other, constative. But what we find in Petrey is a much more balanced position than that held by Kearns. Petrey concludes that “illocutionary force is a combination of language and social practice” (13). While Petrey pays attention to both language and social practice, Kearns only sees the power of context in determining the force of “any utterance” (11). Now, given a sentence “The constitution was enacted in 1980,” no matter whether it appears in a newspaper article or a government decree, it will not be performative, since it merely describes an event of the past in each case. Indeed, the performative power of “The constitution is [hereby] suspended” in the government decree rests both on the linguistic structure and the felicity conditions.

       Now, if “the right context can cause almost any text to be taken as narrative,” the distinction between narrative and non-narrative would in itself become meaningless since we will be left with only one category of texts in “the right context.” And if context can make “any text” or all texts just one category, all generic distinctions will become irrelevant, and the investigation of texts according to generic conventions will be out of place. As touched on above, the investigation of speech-act theorists is based on the distinction between different genres of discourse (e.g. a newspaper article vs. a government decree; a traditional novel vs. a new novel) or different situational contexts (e.g. a courtroom trial vs. a televising report). They take for granted the generic identity of a text, and proceed to investigate the generic conventions operating in the interpretive process, hence displaying the weak-contextualist position.

Kearns states, “Speech-act theory in fact provides the basis for my strong-contextualist position, hence for a truly ‘rhetorical’ narratology. This theory has been defined as nothing less than ‘an account of the conditions of intelligibility, of what it means to mean in a community, of the procedures which must be instituted before one can even be said to be understood’ (Fish 1024).” (Kearns 10) But precisely because speech-act theorists lay emphasis on “conditions of intelligibility,” they pay particular attention to a text’s own generic identity and the conventions of the genre to which the text belongs. Pratt, for instance, asserts, “What I do claim is that regardless of what form the fictional utterance actually takes in a novel, the fact that the text is a novel automatically entitles the reader to bring these rules to bear on the fictional speech act” (206). If, say, Robbe-Grillet’s Dans le labyrinthe is not interpreted as a new novel, but as a non-fictional daily narrative under the influence of an extratextual context, the book surely will not make sense. It should have become clear that, with the strong position that grants the context all determining power, the investigation cannot be truly rhetorical, since it allows texts to be interpreted with “mistaken” conventions and even go so far as to obliterate generic distinctions. Moreover, the investigation cannot be truly narratological, since it ignores the nature of the text itself, allowing, say, a non-narrative to be interpreted as narrative, or vice versa, under the influence of context.

      It is, however, significant to note that Kearns Rhetorical Narratology presents a remarkable synthesis of various approaches, extremely rich in references and comprehensive in coverage of the related fields. This open-minded incorporation and synthesis of various theories function to redress the one-sidedness of Kearns’ own theoretical model. For instance, Kearns’ incorporating Bleich’s view on gender and reading and Lanser’s view on narrating voice etc. takes him beyond the “situational context” into the sociohistorical context in his narratological criticism. Moreover, formal narrative poetics as a technical basis survives. Classical narratological models even dominate the two sections “Temporal Structure of Narrating” (140-52) and “Representing Speech” (152-61). Kearns makes the following comment on Genette’s taxonomy of narrative anachronies:

On the one hand, a narrative’s deviations from a strictly linear ordering of events are true to the human experience of time, and the different types of deviations (such as flashback and flash-forward, in ordinary terms) affect readers differently. On the other hand, this taxonomy says nothing about how important anachronies may be in a particular novel, how they may operate in the time-bound process of reading. To put the point in a practical light, students can be taught the scheme, just as they can be taught the main types of poetic feet. But they must also be led to understand that no “prolepsis” (Genette’s term for flash-forward) is important in itself, that the personal, textual, rhetorical, and cultural contexts have much to do with whatever value the element carries. (5, my emphasis)

From this observation, we can see clearly the difference between formal narrative poetics and contextual criticism. The former is concerned with the basic schemes of narrative fiction as a genre, while the latter explicates given works in the relevant “personal, textual, rhetorical, and cultural contexts.” From Kearns’ observation, it can also be inferred that different narrative devices, such as flashback and flash-forward, have both varied contextualized significance (when used in specific narratives) and shared generic functions, the latter being able to “affect readers differently” in a decontextualized way. The investigation of the effect of a narrative device in a specific narrative must take account of both kinds of meaning. Formal narrative poetics provides classifications of narrative devices and discusses their generic functions, thus offering tools helpful for answering the relevant questions about narrative in various kinds of context. Seen in this light, it is not surprising that despite Kearns’ theoretical exclusion of formal narrative poetics from a strong-contextualist position, classic narrative poetics constitutes the most important technical basis for his rhetorical investigation.

Conclusion

       It should have become clear that formal narrative poetics has been existing and developing within various kinds of contextual narratologies, where there can be found a triple mutually-benefiting relationship: (1) new decontextualized structural models are developed for contextual criticism; (2) the theoretical contributions may both depend upon and expand classical narrative poetics; (3) contextual criticisms draw on classical structural tools, which, in turn, helps classical narrative poetics to gain current relevance. Clearly, there is no real ground for the antagonism between formal narrative poetics and contextual narratologies (see, for instance, the debate between Diengott and Lanser in Style (1988)).

As an important technical basis for various kinds of contextulist narratological analyses, formal narrative poetics needs to be continuously and rigorously improved and developed. As indicated above, existing narrative poetics is not free from limitations, calling for continuous revision and enrichment (see also Shen “Narrative”; “Defense”; “Difference”), for the use of new tools (see for instance, Ryan Possible; “Cyberage”), and for further extension into new areas (see also Cuddy-Keane; Richardson “Beyond”). So long as existing structural models leave room for improvement, so long as the generic structures of existing narratives in various genres and media are not exhausted by description, and so long as new structures and techniques emerge in future narratives providing room for further narratological classification or description, formal narrative poetics can be “a mode of theorizing that is open, dynamic, neverending” (Rimmon-Kenan, Narrative 48).

      Talking about the situation of narratology at the end of the last century, Mieke Bal depicts a rather deplorable picture: “Text-grammars have ceased to appear, formalist models are deemed irrelevant, and while some hold on to early structuralist distinctions, many of those who discussed the criteria by which to spot Free Indirect Discourse in 1979 moved on, and practise analysis rather than worrying about how to do it” (Narratology 13). Indeed, many narratologists have turned to narratological criticism, but as Bal puts it, “application may imply an unwarranted acceptance of imperfect theories” (ibid.). While in some countries, formal narrative poetics as a domain “continues to be practised and amplified by a relatively limited number of specialists” (Rimmon-Kenan, Narrative 135), in North America at least, formal narrative poetics for the past 15 years or so has been reduced to a position of being survivable only under the “disguise” of part of contextualist or postclassical narratologies. This apparently has hindered its development. It should be stressed that to promote the development of formal narrative poetics—both within contextual narratologies and as another domain (as an increasing repertory of tools for contextual criticism)—is very significant for the development of contextual narratologies. It should also have become clear that asking contextual approaches to stay away from narratolgy (Diengott) is to shut the door to beneficial, invigorating forces and resources, indeed to the very “savior” of narratology at least with reference to North America. For the sake of the development of the field of narratology, it is high time to clarify the different relations the different kinds of narratological investigation bear to reader and context, to distinguish between generic function and contextualized significance, and to see the relation between contextual narratologies and formal narrative poetics not as one of mutual exclusion, but as one of mutual nourishment.

ENDNOTES

1.  David Darby’s essay and the debate it generated in Poetics Today 24.3 (2003) focus on the relation between the situation in Germany and that in North America.

2.     However, the “authorial voice” in modern times no longer enjoyed the same degree of its traditional authority, and in postmodern times, may even be found out of place, but it is still relatively more authoritative than the “personal voice” once it appears in narratives – unless it is parodied or used in some other ironic way.

3.   The situation is very different in China. The late 1970s and early 1980s saw a sudden rush into China of various schools of Western literary theory and criticism. Having been subjected to political criticism for decades, many Chinese scholars became particularly interested in text-oriented schools. Thus when the American academic world was trying to move away from the limitations of formalist criticism towards a better understanding of the connections between literature and society, history, and ideology, many scholars in China moved away from political criticism towards the study of form and aesthetics. Around 1980, extrinsic criticism was even temporarily excluded in China. But fortunately the situation soon became more tolerant and Chinese scholars became increasingly aware of the limitations of formalist or intrinsic criticism. Contextual narratologies and formal narrative poetics have been enjoying mutual development recently. If China around 1980 and North America since 1980s constitute two contrastive poles, other countries may fall somewhere in between.

4.       For further discussions on the different reading positions, see Booth; Rabinowitz 1987; Phelan 1989, 1996, 2005; Kearns; Herman 2002.

5.             Even in the case of “narrativization” (“a reading strategy that naturalizes texts by recourse to narrative schemata” (Fludernik, Towards 34)), textual features may still have a vital role to play. Monika Fludernik offers the following two examples of narrativization: Alain Robbe-Grillet’s La Jalousie has frequently been narrativized as the jealous husband’s observations of his wife through the window shutters; and the “very term of ‘camera-eye’ technique” “betrays another such narrativization which attempts to correlate the text with a frame from recognizable experience.” (Towards 46) Now, in La Jalousie, the text frequently suggests that the angle of vision comes from behind the window shutters (apparently, the author wrote in such a way as to invite readers to interpret the text as such). As for “camera-eye,” this metaphorical term was borrowed by Norman Friedman from the narrator’s opening statement in Isherwood’s Goodbye to Berlin, “I am a camera with its shutter open, quite passive, recording, not thinking” (quoted in Friedman 1179). It is a term that is only applicable to texts whose “aim is to transmit, without apparent selection or arrangement, a ‘slice of life’ as it passes before the recording medium” (Friedman 1179), and not applicable to any other kind of texts.

6.             Kearns asserts that “narrativity, like fictionality, is a function of context” (35), as opposed to text. But I would argue that while fictionality often does not have to do with textual features, textual elements have a much more important role to play in determining narrativity (even in the former case, context alone cannot determine fictionality: a real telephone directory is a real telephone directory even though a given context leads the reader to treat it as fictional). As distinct from fictionality, narrativity, in general, should be conceived “as both textually exterior and interior” (Prince, “Remarks” 104). Now, given “two plus two equals four” or “San Franscisco is a beautiful city” or “trees must be preserved for environmental protection”, I’m sure no one will claim that any of those sentences can function as narrative in a proper context (see also note 5).     

 WORKS CITED

Bal, Mieke. Death and Dissymetry. Chicago: U of Chicago P, 1988.

---. “The Point of Narratology.” Poetics Today 11 (1990): 727-852.

---. Narratology. Trans. Christine van Boheemen. Toronto: U of Toronto P, 2^nd edition 1997.

Bleich, David. “Gender Interests in Reading and Language.” Gender and Reading. Ed. Elizabeth Flynn and Patrocinio P. Schweickart. Baltimore: Johns Hopkins UP, 1986. 234-66.

Booth, Wayne C. The Rhetoric of Fiction. Harmondsworth: Penguin Books, 2^nd edition 1983.

Bortolussi, Marisa and Peter Dixon. Psychonarratology. Cambridge: Cambridge UP, 2003.

Case, Alison. Plotting Women. Charlottesville: UP of Virginia, 1999.

---. “Gender and History in Narrative Theory: The Problem of Retrospective Distance in David Copperfield and Bleak House.” A Companion to Narrative Theory. Ed. James Phelan and Peter J. Rabinowitz. Oxford: Blackwell, 2005. 312-21.

Chatman, Seymour. Story and Discours. Ithaca: Cornell UP, 1978.

---. Coming to Terms. Ithaca: Cornell UP, 1990.

---. “What Can We Learn from Contextualist Narratology?” Poetics Today 11 (1990): 309-328.

Cuddy-Keane, Melba. “Modernist Soundscapes and the Intelligent Ear: An Approach to Narrative Through Auditory Perception.” A Companion to Narrative Theory. Ed. James Phelan and Peter J. Rabinowitz. Oxford: Blackwell, 2005. 382-398.

Culler, Jonathan. Structuralist Poetics. London: Routledge & Kegan Paul, 1975.

Currie, Mark. Postmodern Narrative Theory. New York: St. Martin, 1998.

Darby, David. “Form and Context: An Essay in the History of Narratology.” Poetics Today 22 (2001): 829-852.

Derrida, Jacques. Positions. Paris: Minuit, 1972.

Diengott, Nilli. “Narratology and Feminism.” Style 22 (1988): 42-51.

Fetterley, Judith. The Resisting Reader. Bloomington: Indiana UP, 1978.

Fish, Stanley. “How to Do Things with Austin and Searle: Speech Act Theory and Literary Criticism.” Modern Language Notes 91 (1977): 983-1025.

---. Is There a Text in this Class? Cambridge, MA: Harvard UP, 1980.

Fludernik, Monika. Towards a “Natural” Narratology. London: Routledge, 1996.

---. “Histories of Narrative Theory (II): From Structuralism to the Present.” A Companion to Narrative Theory. Ed. James Phelan and Peter J. Rabinowitz. Oxford: Blackwell, 2005. 36-59.

Friedman, Norman. “Point of View in Fiction: The Development of a Critical Concept.” PMLA 70 (1955): 1160-84.

Genette, Gérard. Narrative Discourse. Trans. Jane E. Lewin. Ithaca: Cornell UP, 1980.

Herman, David. “Introduction: Narratologies.” Narratologies. Ed. David Herman. Columbus: Ohio State UP, 1999. 1-30.

---. Story Logic. Lincoln: U of Nebraska P, 2002.

---, ed. Narrative Theory and the Cognitive Sciences. Stanford: CSLI, 2003.

Herman, David, et al. “Introduction.” The Routldge Encyclopedia of Narrative Theory. Ed. David Herman et al. London: Routledge, 2005. ix-xi.

Jahn, Manfred. “‘Speak, friend, and enter’: Garden Paths, Artificial Intelligence, and Cognitive Narratology.” Narratologies. Ed. David Herman. Columbus: Ohio State UP, 1999. 167-94.

Kafalenos, Emma. “Not (Yet) Knowing: Epistemological Effects of Deferred and Suppressed Information in Narrative.” Narratologies. Ed. David Herman. Columbus: Ohio State UP, 1999.

Kearns, Micheal. Rhetorical Narratology. Lincoln: U of Nebraska P, 1999.

Lanser, Susan Sniader. The Narrative Act. Princeton, New Jersey: Princeton UP, 1981.

---. “Toward a Feminist Narratology.” Style 20 (1986): 341-363. Reprinted in Feminisms: An Anthology. Ed. Robyn R. Warhol and Diane Price Herndl. New Brunswick: Rutgers UP, 1991. 610-629.

---. “Shifting the Paradigm: Feminism and Narraology.” Style 22 (1988): 52-60.

---. Fictions of Authority. Ithaca: Cornell UP, 1992.

---. “Sexing the Narrative: Propriety, Desire, and the Engendering of Narratology.” Narrative 3 (1995): 85-94.

Margolin, Uri. “Narrative and Indexicality: A Tentative Framework.” Journal of Literary Semantics 13 (1984): 181-204.

Mezei, Kathy. “Introduction: Contextualizing Feminist Narratology.” Ambiguous Discourse. Ed. Kathy Mezei. Chapel Hill: U of North Carolina P, 1996. 1-20.

---, ed. Ambiguous Discourse. Chapel Hill: U of North Carolina P, 1996.

Nunning, Ansgar. “Towards a Cultural and Historical Narratology: A Survey of Diachronic Approaches, Concepts, and Research Projects.” Anglistentag 1999 Mainz Proceedings. Trier: Wissenschaftlicher Verlag, 1999. 345-73.

Onega, Susana and Jose Angel Garcia Landa. Narratology. London: Longman, 1996.

Page, Ruth E. “Feminist Narratology? Literary and Linguistic Perspectives on Gender and Narrativity.” Language and Literature 12 (2003): 43-56.

Petrey, Sandy. Speech Acts and Literary Theory. London: Routledge, 1990.

Phelan, James. Reading People, Reading Plots. Chicago: U of Chicago P, 1989.

---. Narrative as Rhetoric. Columbus: Ohio State UP, 1996.

---. Living to Tell about It. Ithaca: Cornell UP, 2005.

Phelan, James and Peter J. Rabinowitz. “Introduction: Tradition and Innovation in Contemporary Narrative Theory.” A Companion to Narrative Theory. Ed. James Phelan and Peter J. Rabinowitz. Oxford: Blackwell, 2005. 1-16.

Pratt, Mary Louise. Towards a Speech Act Theory of Literary Discourse. Bloomington: Indiana UP, 1977.

Prince, Gerald. Dictionary of Narrayology. Lincoln: U of Nebraska Press, 1987.

---. “On Narratology: Criteria, Corpus, Context.” Narrative 3 (1995): 73-84.

---. “Remarks on Narrativity.” Perspectives on Narratology. Ed. Claes Wahlin. Frankfurt: Peter Lang, 1996. 95-106.

Rabinowitz, Peter J. “Truth in Fiction: A Reexamination of Audiences.” Critical Inquiry 4 (1977): 121-41.

---. Before Reading. Ithaca: Cornell UP, 1987.

Richardson, Brian. “Beyond Story and Discourse: Narrative Time in Postmodern and Nonmimetic Fiction.” Narrative Dynamics. Ed. Brian Richardson. Columbus: Ohio State UP, 2002. 47-64.

Rimmon-Kenan, Shlomith. A Glance Beyond Doubt. Columbus: Ohio State UP, 1996.

---. Narrative Fiction: Contemporary Poetics. London: Routledge, 2^nd edition 2002.

Robinson, Sally. Gender and Self-Representation in Contemporary Women’s Fiction. Albany: State U of New York P, 1991.

Ryan, Marie-Laure. Possible Worlds, Artificial Intelligence, and Narrative Theory. Bloomington: Indiana UP, 1991.

---. “Cyberage Narratology: Computers, Metaphors, and Narrative.” Narratologies. Ed. David Herman. Columbus: Ohio State UP, 1999. 113-41.

---. “Cognitive Maps and the Construction of Narrative Space.” Narrative Theory and the Cognitive Sciences. Ed. David Herman. Stanford: CSLI, 2003. 214-242.

Saussure, Ferdinand de. Course in General Linguistics. Trans. Wade Baskin. London: Philosophical Library Inc., 1960.

Schweickart, Patrocinio. “Reading Ourselves: Toward a Feminist Theory of Reading.” Gender and Reading. Ed. Eliazabeth Flynn and Patrocinio Schweichart. Baltimore: The Johns Hopkins UP, 1986. 31-62.

Shen, Dan. “Stylistics, Objectivity, and Convention.” Poetics 17 (1988): 221-238.

---. “Narrative, Reality, and Narrator as Construct: Reflections on Genette’s ‘Narrating’.” Narrative 9 (2001): 123-129.

---. “Defence and Challenge: Reflections on the Relation between Story and Discourse.” Narrative 10 (2002): 222-243.

---. “The Future of Literary Theories: Exclusion, Complementarity, Pluralism.” Ariel 33 (2002): 159-182.

---. “Difference Behind Similarity: Focalization in Third-Person Center of Consciousness and First-Person Retrospective Narration.” Acts of Narrative. Ed. Carol Jacobs and Henry Sussman. Stanford: Stanford UP, 2003. 81-92.

Sternberg, Meir. “Universals of Narrative and Their Cognitivist Fortunes (I).” Poetics Today 24 (2003): 297-395.

Warhol, Robyn. Gendered Interventions. New Brunswick: Rutgers UP, 1989.

---. “The Look, the Body, and the Heroine of Persuasion.” Ambiguous Discourse. Ed. Kathy Mezei. Chapel Hill: U of North Carolina P, 1996. 21-39.

---. “Neonarrative; or, How to Render the Unnarratable in Realist Fiction and Contemporary Film.” A Companion to Narrative Theory. Ed. James Phelan and Peter J. Rabinowitz. Oxford: Blackwell, 2005. 220-31.

   语境叙事学与形式叙事学缘何相互依存[*]