What is Exposition? An Essay in Temporal Delimitation

What is Exposition? An Essay in Temporal Delimitation

MEIR STERNBERG

In line with Wolfgang Iser, Umberto Eco and Horst Ruthrof, Meir Sternberg offers a reader-response approach to narrative. He defines narrative as a system of 'gap filling' and hypothesis testing and contends that the textual structure must be studied as it is dynamically constructed by the reader. The endurance of first impressions is a law of human perception. First impressions condition further perceptions. Sternberg therefore proposes the elaboration of a scale of rhetorical effects according to 'the rise and fall of first impressions' and a classification of retardatory structures in the presentation of narrative information. He likewise contends that literary texts are the result of the selection and combination of motifs and that the temporal distortion of the chronological order of events is an indication of artistic purpose.

In this excerpt from chapter 1 of his book Expositional Modes and Temporal Ordering in Fiction, Sternberg attempts to define exposition functionally and transgenerically. He contends that the Russian formalist distinction between fabula and siuzhet is not equivalent but complementary to E. M. Forster's notions of 'story' and 'plot'. He defines exposition as the beginning of the fabula; that is, as the first part of the chronologically ordered sequence of motifs as reconstructed by the reader, which may or may not coincide with the beginning of the siuzhet. Sternberg further differentiates between represented time and representational time ( Genette's story time and narrative time). He argues that every work establishes its own time-norm and that there is a logical correlation between the amount of time devoted to an element and the degree of its aesthetic relevance or centrality. For Sternberg, the closer the two times become, the more aesthetically relevant is the scene.

As the whole of anything is never told, the writer of fiction is necessarily confined to presenting his characters in action within the

limits of a certain fictive period of time. It is thus unavoidable that he should intersect the lives of his dramatis personae at a given hour. His problem is only to decide which hour it shall be and in what situation they shall be discovered. [ . . . ]

It is the function of the exposition to introduce the reader into an unfamiliar world, the fictive world of the story, by providing him with the general and specific antecedents indispensable to the understanding of what happens in it. There are some pieces of information, varying in number and nature from one work to another, that the reader cannot do without. He must usually be informed of the time and place of the action; of the nature of the fictive world peculiar to the work or, in other words, of the canons of probability operating in it; of the history, appearance, traits and habitual behavior of the dramatis personae; and of the relations between them.

In some instances it may indeed seem (though I shall argue this is not the case) that a certain amount of prior information -- about the characters and the fictive world -- that is not fully contained in the work itself may be assumed beforehand. In Greek drama, for example, the dramatists, restricted to a well-defined field of material, told and retold myths with which their audience was familiar. Whenever the narrative materials are derived from history, it may likewise seem that the communication of at least part of the expositional information may be dispensed with on the assumption that the author takes for granted his reader's possession of a certain amount of common knowledge. [ . . . ]

Even a number of modern writers may seem to share the expositional privileges or exemptions of their ancient predecessors. I am referring especially to novelists celebrated for their progressive creation of some private, full-fledged fictive world -- Trollope's Barchester, Balzac's nineteenth-century France, or Faulkner's Yoknapatawpha County -- repeatedly carrying over not only settings but whole casts of characters and clusters of incidents from one work to another of the same cycle. But the same may be true of any series of works, notably detective stories, in which at least one central character recurs (e.g., Agatha Christie's Hercule Poirot), even though the setting of the fictive world varies.

Many critics work on the implicit (and sometimes even explicit) assumption that in all these cases at least part of the expositional antecedents may indeed be taken as known or obtained by the reader outside the limits of the single work, particularly with reference to different stories of the same cycle, which they regard as a single unit. A close examination of the literary evidence, however, indicates that this assumption is untenable. In their contempt for the fatal futility of Fact, writers usually have no scruples about supplementing, modifying, or even distorting historical evidence or tradition to suit their artistic purposes. Shakespeare is notorious for the free use he made of his sources. In Julius Caesar, he drastically both simplifies and complicates the history of the two years between Caesar's triumphant return to Rome and the decisive battle at Philippi. ¹ [ . . . ] Still more audaciously, a popular historical tradition may be initially embodied in the work only to be demolished or reversed at a later stage, as is the case with the gruesome figure of Richard III in Josephine Tey's The Daughter of Time.

When a character or a situation is carried over from one work to another, the writer feels no less free to introduce in them any changes dictated by the distinctive artistic conception of the new work. [ . . . ] As Malcolm Cowley himself admits, 'as one book leads into another, Faulkner sometimes falls into inconsistencies of detail. . . . Henry Armstid is a likable figure in As I Lay Dying and Light in August; in The Hamlet he is mean and half-demented'; and so on. ² [ . . . ] Whether we are to shrug such changes off as 'inconsequential errors,' ³ as Cowley does, or, as can easily be established, take them to be deliberate and revealing deviations from previous thematic and structural conceptions, it is evident that they constitute or call for new expositional material. [ . . . ]

Moreover, notwithstanding general ex cathedra declarations to the contrary, writers as a rule take the necessary precautions to render each of their works as expositionally autonomous as possible, even when the carrying over of characters and fictive world involves no divergence from previous conceptions. In the second chapter of Trollope Barchester Towers, for instance, the narrator informs the reader that 'it is hardly necessary that [he] should here give to the public any lengthened biography of Mr. Harding up to the period of the commencement of this tale.' [ . . . ] Although Trollope ostensibly professes to assume that Mr. Harding's ordeal, formerly narrated in The Warden, must by now be a matter of common knowledge, he in fact cunningly recapitulates the occurrences expositionally. relevant to Barchester Towers. What was there the core of the action proper is here telescoped into a few passages; and some additional sentences then bring the account up to date and effect the necessary transition to 'the commencement of this tale.' [ . . . ]

In conclusion, [ . . . ] the limits of a literary unit cannot be fixed a priori but are dynamic in that they vary according to the kind of questions the critic poses. [ . . . ] Exposition, therefore, can never be dispensed with with impunity; and the peculiar problems it raises must be confronted and solved by every writer, in every work afresh.

The location of exposition

So far I have discussed the distinctive function of the exposition. The question arises, however, whether, bearing in mind this function, we can point to any specific part or parts of a narrative work (or any literary text subsuming a narrative sequence) that can be called 'the exposition.' What is the location of the expositional sections or elements? Is it fixed or variable? And finally, how are the expositional sections or elements to be distinguished from the nonexpositional?

The most detailed and widely accepted theory of exposition is the time-honored view first proposed by Gustav Freytag, whose scheme of dramatic structure includes exposition as an integral part: 'The drama possesses . . . a pyramidal structure. It arises from the introduction with the entrance of the exciting forces to the climax, and falls from there to the catastrophe. Between these three parts lie (the parts of the rise and the fall.' ⁴ An important dramatic effect called the exciting moment or force 'stands between the introduction and the rise': 'the beginning of the excited action (or complication) occurs at a point where, in the soul of the hero, there arises a feeling or volition which becomes the occasion of what follows; or where the counterplay resolves to use its lever to set the hero in motion. . . . In Julius Caesar this impelling force is the thought of killing Caesar, which, by the conversation with Cassius, gradually becomes fixed in the soul of Brutus' (pp. 115, 121 ). [ . . . ]

Freytag's conception of exposition, however, plausible and tidylooking as it is, seems to me untenable. Its fatal weakness consists not so much in its limited range of applicability as in its internal inconsistency and its failure to stand up to the facts even when tested against works that are constructed 'pyramidally.' If the function of the exposition is, in Freytag's own words, 'to explain the place and time of the action, the nationality and life relations of the hero' (pp. 117 -18), it is hardly possible to prescribe or to determine a priori that all authors must invariably choose to locate the expositional information within the first act or before the 'rising action.' And indeed, writers seldom impose on themselves any limitations of this kind. In Ibsen Ghosts, for instance, the exposition is distributed throughout, and new vital facts concerning the past of the agents keep cropping up as late as the last act. But the fallacy can be demonstrated even with reference to the plays. Freytag himself cites in illustration of his theory. In Julius Caesar, he maintains, the exciting force 'is the thought of killing Caesar, which, by the conversation with Cassius [act 1, scene 2], gradually becomes fixed in the soul of Brutus' (p. 121 ). In fact, the exposition is not concentrated within the limits of the first act, and only a small part of it precedes the impelling moment. Most of the expositional material is widely distributed: one important aspect of Brutus's expositional 'life relations,' his relations with his wife Portia, is 'explained' only in act 2, scene 1, after Brutus has already assumed the leadership of the conspiracy; Caesar's 'life relations' with Calpurnia are dramatically conveyed even later, in act 2, scene 2; while the full disclosure of Antony's relations with Caesar is delayed until his famous soliloquy in act 3, scene 1. These various antecedents ('life relations'), all of them indisputably expositional according to Freytag's definition of the function of exposition, turn out to be as indisputably nonexpositional according to his description of its location; and as Freytag's definition of the function of exposition is basically sound, we must conclude that his prescriptive view of its location must be wrong. [ . . . ]

The weakness of Freytag's theory of exposition stems, in fact, from a major flaw in his general model of structure. Freytag purports to describe the structure of the action as a movement in time, in a definite direction and through definite stages -- in the temporal order in which the reader or audience learns of the developments of the action. But what he really describes is not the movement of the action but the structure of the conflict. He divorces this from the actual temporal movement of the action, presenting a structure that is viewed by the reader only when he retrospectively looks back on the action and rearranges or reassembles it chronologically in his mind. What Freytag and his followers fail to take into account is that the chronological order in which events happen need not necessarily coincide with the order in which they are imparted to the reader. Consequently, the 'absolute,' chronological order of occurrence (in which exposition is indeed preliminary in point of time) does not necessarily correspond with the actual temporal movement or order of presentation of the same events in an actual work, in which expositional information may even be deferred to the last scene or chapter, as it is in Gogol Dead Souls or in the detective story.

In short, as innumerable literary works where the exposition or part of it is either delayed or distributed cannot be fitted into Freytag's Procrustean scheme, his claims about the fixed and static location of exposition must be rejected. The only acceptable theory of exposition will be one flexible enough to hold good equally for all kinds of structure and to cut across the boundaries of genre.

Exposition, fabula and sujet, story and plot

It seems to me possible to define exposition satisfactorily only in terms of fabula, sujet, and scenic norm. The important distinction between fabula and sujet, first proposed by the Russian Formalists, ⁵ is still amenable to further discrimination and development. A narrative work is composed of myriads of motifs, that is, basic and contextually irreducible narrative units. ⁶ Examples of such motifs in The Ambassadors would be 'Strether reached the hotel at Chester,' 'He found himself facing a lady in the hall,' or 'Waymarsh made a sudden dash into a shop.' The fabula of the work is the chronological or chronological-causal sequence into which the reader, progressively and retrospectively, reassembles these motifs; it may thus be viewed as the second-degree 'raw material' (postselected and straightforwardly combined narrative) that the artist compositionally 'deforms' and thus re-contextualizes in constructing his work (mainly by way of temporal displacements, manifold linkage, and perspectival manipulations). The sujet, in contrast, is the actual disposition and articulation of these narrative motifs in the particular finished product, as their order and interrelation, shaping and coloring, was finally decided on by the author. To put it as simply as possible, the fabula involves what happens in the work as (re)arranged in the 'objective' order of occurrence, while the sujet involves what happens in the order, angle, and patterns of presentation actually encountered by the reader. [ . . . ]

Apart from this -- to repair a pronounced bias of the Formalists' -one must also take into account that the fabula is equally amenable to manipulations of point of view, a form of artistic deformation and re-contextualization that frequently coincides with and sometimes even accounts for temporal displacements. The author can postulate an omniscient narrator, or compose an epistolary story, or employ any of the characters as the narrator, or record the action as it passes through the consciousness of any or all of them; in each case the temporal order of the motifs, their combination, significance, weight, and coloring will vary. Henry James used to say that there are five million ways of telling a story. He meant, of course, that out of a given, basically similar fabula, five million sujets can be molded, each with its own temporal structure and narrative strategy and consequently with its own peculiar effect on the reader. [ . . . ]

It is accordingly in terms of the distinction between fabula and sujet that I can now redefine my main objection to Freytag's theory -its failure to differentiate the absolute dynamics of the causally

propelled action from the variable dynamics of the reading-process. To assert that the first act of any play (or the first few chapters of a novel) contains the exposition is to confuse the beginning of the sujet and that of the fabula. The exposition always constitutes the beginning of the fabula, the first part of the chronologically ordered sequence of motifs as reconstructed by the reader; but it is not necessarily located at the beginning of the sujet. The two beginnings coincide and overlap only when the author presents his tale in a straight chronological sequence (as happens, more or less, in the Book of Job, or in the Laxdaela Saga, or in James Washington Square). The author, however, may as legitimately choose to plunge in medias res or to distribute the expositional material throughout the work; and in these cases, though the exposition is still located at the beginning of the fabula, its position in the order actually devised to present the motifs to the reader radically varies.

Represented time and representational time: the quantitative indicator

So far I have defined exposition as the 'beginning' or 'first part' of the fabula. This definition, however, though it firmly establishes the expositional terminus a quo and though it flexibly covers the innumerable possibilities of combining and ordering a given number of motifs, may still be regarded as seriously incomplete unless we can determine exactly up to what point in the fabula the motifs are expositional. To discover this elusive line of demarcation, we must first consider more of the time values of fiction (time as a dimension, object and indicator of artistic selection as well as of combination and ordering) and the important role they play in guiding the reader's interpretation of the work.

Narrative presents characters in action during a certain fictive period of time. As a rule, however, one finds that the author has not treated the whole of the fictive period in the life of the characters with the same degree of attention. This period falls naturally or is artificially divided into different subperiods, stages, or time-sections. Some of these are rendered at great length, some galloped through or rapidly summarized, some dismissed with a perfunctory sentence or two, while others are even passed over unmentioned. Even within the, framework of a single work, therefore, we generally discover different ratios of represented time (i.e., the duration of a projected period in the life of the characters) to representational time (i.e., the time that it takes the reader, by the clock, to peruse that part of the text projecting this fictive period).

The differentiation between what I call representational and represented time dates back, in fact, to the Renaissance and the NeoClassical age, during which it was exclusively employed as a normative tool for checking the adherence of dramatists to the so called Aristotelian unity of time. Castelvetro, for example, distinguishes 'perceptible time' from 'intellectual time', and Dryden denounces the practice of 'mak[ing] too great a disproportion betwixt the imaginary time of the play, and the real time of its representation.' The concern with time-ratios has, moreover, been revived in modern criticism. Various German scholars distinguish 'erzhlte Zeit' from 'erzhlzeit'; and A. A. Mendilow elaborates a similar distinction between 'the chronological duration of the novel' and 'the chronological duration of the reading.' I believe, however, that these various pairs of terms have not been sufficiently exploited. They have traditionally been used mainly to indicate the ratio between the representational and the represented time of the work as a whole -- 'the time it takes to read a novel' as against 'the length of the time covered by the content of the novel' -- and also the disparities between different works in this regard, but less often to investigate the variations in time-ratios within a single work. And even when such variations have been pointed out, this has usually been done in order to discuss their implications for the work's tempo or its narrative rhythm. I certainly agree that a comparison of the time-ratios in different works may yield highly significant results, some of which will be referred to below. [ . . . ] It is, however, at least equally important to trace the variations in time-ratio within the limits of a single work as well; for these variations not only lead the reader to various 'formal' conclusions (as to tempo) but at the same time play a central role in the interpretation of the text.

The reader is always confronted and frequently baffled by such questions as, Who is the protagonist or center of interest in the work? What is the relative importance of the various characters, incidents and themes? And how do they combine with the center? He is obliged to pose and answer dozens of questions of this sort if he is to construct, or reconstruct, the work's structure and hierarchy of meaning and to compose a fully integrated picture of its art. However, these questions are never settled explicitly and satisfactorily by the text itself, even when overt rhetoric is employed -- not even when the author refers to one of the characters as 'my hero' or openly calls the reader's attention to the role played by a certain incident or agent. The reader is therefore forced to follow the multifarious implications of the text (its dramatized rhetoric) as to its peculiar principles of selection and combination in order to work out adequate answers. The quantitative indicator, revealing the principles of selection operating in the text, forms one of the reader's indispensable guides in the process of interpretation in that it helps him to determine the text's general tendency ('intention') and its particular structure of meaning. For owing to the selectivity of art, there is a logical correlation between the amount of space devoted to an element and the degree of its aesthetic relevance or centrality, so that there is a good prima facie case for inferring the latter from the former.

As the variations in time-ratios form one manifestation of the quantitative indicator, it can be determined that, mutatis mutandis, the time-ratio of a narrative time-span or event generally stands in direct proportion to its contextual relevance: one whose representational time approximates its represented time is implied to be more central to the work in question than another in which these two time factors are incommensurate. [ . . . ]

It is quite understandable that different writers, each with his own conception of life and poetics of art, should differ as to what fields of material merit (thorough) treatment. But the reader finds himself in an altogether different situation. Qua reader, he has no private artistic axe to grind. His only business is to endeavor to grasp the nature and functions of the compositional principles operating in the text, so that he may comprehend as fully as possible its structure of meaning. Having this in view, he cannot apply to the work any scale of intrinsic interest (including his own), because there is not a single one that is universally valid. He must, therefore, measure the value of narrative elements in terms of contextual significance, largely suggested by the quantitative indicator. ⁷ [ . . . ] Laurence Sterne, for example, demonstrates his acute awareness of the functionality of his seemingly bizarre selective procedure when he claims that 'the happiness of the Cervantic humour arises from this very thing -- of describing silly and trifling Events, with the Circumstancial Pomp of great Ones.'

Fielding himself was, in fact, well aware of the value-determining aspect of the temporal variations, though in his polemical impetus against what seemed to him Richardson's petty psychological preoccupations he at times tended to overstate his plea to the contrary. Regardless of the overt motivation of his selective decisions, they are actually based not on the ostensible criterion of intrinsic interest versus dullness but on that of artistic relevance versus irrelevance, as he himself is driven to admit openly immediately after the muff incident in Tom Jones:

Though this incident will probably appear of little consequence to many of our readers, yet trifling as it was, it had so violent an effect on poor Jones that we thought it our duty to relate it. In. reality, there are many little circumstances too often omitted by injudicious historians, from which events of the utmost importance arise. [ . . . ]

Quantitative indicator, scenic norm, and fictive present

The quantitative indicator is also an indispensable factor in the delimitation of the exposition, especially in determining the precise temporal point in the fabula which marks the end of the exposition.

As argued above, the literary artist exploits the possibilities of varying the time-ratios in order to throw the contextual centrality of certain fictive periods into high relief against the background of other periods belonging to the total span of the sujet. It is thus the approximation of representational to represented time that draws the reader's attention to some subperiods constituting 'discriminated occasions' in the [Jamesian] sense of the word. And vice versa: the very disparity between the different time-ratios (and the greater it is the more conspicuously significant it becomes) suggests that the cursorily treated time-sections are nondiscriminated because they are meant to occupy but a relatively minor position in the particular structure of meaning established by the work.

Moreover, in most fictional (and dramatic) works we find not only variations but also a basic similarity between the time-ratios of the various scenes or discriminated occasions. Every narrative establishes a certain scenic time-norm of its own. This norm may, of course, vary from one writer, and even from one work, to another. And even within a single work certain scenes may turn out to deviate from the basic time-norm established by the majority of the discriminated occasions. But such deviations (say, a ratio of 2:3 or even 1:5 where the norm is 1:2), which may indeed appear considerable when examined in isolation, generally prove insignificant when considered, as they must be, in the context of the whole work -- in the light of the nonscenic as well as the scenic time-ratios. [ . . . ]

Since every work does establish a scenic norm and since the scenic treatment accorded to a fictive time-section underscores its high aesthetic importance, the first scene in every work naturally assumes a special conspicuousness and significance. The author's finding it to be the first time-section that is 'of consequence enough' to deserve full scenic treatment turns it, implicitly but clearly, into a conspicuous signpost, signifying that this is precisely the point in time that the author has decided, for whatever reason, to make the reader regard as the beginning of the action proper. That is, the text suggests, why this 'occasion' is the first to have been so 'discriminated.' [ . . . ] If, therefore, the first discriminated occasion is the beginning of what Trollope happily calls 'the real kernel of [the] story,' ⁸ it follows that any motif that antedates it in time (i.e., precedes it in the fabula) is expositional -- irrespective of its position in the sujet.

The expositional material, always antedating the first scene, may correspondingly precede it in point of its actual position in the sujet. In this case, the large disparity between the time-ratio of the expositional part and that of the opening scene (a disparity concomitant with several other indicators, to be discussed) lays bare the preparatory nature of whatever precedes the temporal signpost. The communication of the expositional material, however, may also be delayed, so that it will succeed the first discriminated occasion in its actual ordering. In this case, the expositional information will retrospectively throw light on it, that is, enrich, modify or even drastically change the reader's understanding of it; for, within the sharply circumscribed, enclosed world of the literary text, almost every motif or occurrence antedating another tends to illuminate it in some way, no matter what their order of presentation in the sujet. The point marking the end of the exposition in the fabula thus coincides with that point in time which marks the beginning of the fictive present in the sujet -- the beginning of the first time-section that the work considers important enough to be worthy of such full treatment as will involve, according to the contextual scenic norm, a close approximation or correspondence between its representational time and the clock-marked time we employ in everyday life.

It will be noticed that I dissociate my use of the term fictive present from any dependence on dramatic or fictive illusion, with which it is usually thought to be interchangeable. [ . . . ]

If we grasp 'fictive present' as a descriptive metaphor denoting an indisputably objective ratio between representational time and represented time, a ratio that involves an approximation of the two times; and if this approximation is interpreted as aiming to achieve (to adopt Mendilow's own phrase in another context) a 'closer correspondence between the pace of living . . . and [the] depiction of it', ⁹ then we shall be able to account for the possibility that a temporal transfer takes place at such a point by referring to the objective compositional elements that may produce it. [ . . . ]

Notes