Notes on Helmut Bonheim's The Narrative Modes: Techniques of the Short Story (Cambridge: D.S. Brewer, 1982). Notes taken by JAGL c. 1986.
Helmut Bonheim
The Narrative Modes: Techniques of the Short Story
Chapter I - Theories of Narrative Modes
1- "Even the shortest of story forms, the anecdote, tends to use all of the chief modes of narrative": (1) description, (2) report, (3) speech, (4) comment.
A Brief History of Modes
Plato, Aristotle... Scott (Waverley, ch. 19). The W's of journalism: "who, when, where", description; "what", report and speech; "why" comment. A conventional sequence. Setting: concrete or abstract, etc. Need for systematization becomes evident in the 70s.
German studies (ignored abroad)
- Petsch (1934), "epische Grundformen": 1) Darstellung, 2) Bericht, 3) Beschreibung, 4) Betrachtung, 5) Bild, 6) Szene, 7) Dialog. (1, 2, 5, 6 are subdivisions of Report).
- Koskimies (1935) adds tableau (dialogue) to this.
- Kayser, Lämmert, elaborate upon the basic modes. Lämmert adds the criterion of time. For Lämmert, Holdheim and Lodge, description is timeless (hard to accept in practice).
The model of four modes recognizes overlappings and the possibility of subdivisions.
- [Uffe] "Hansen (1975) includes such modes as indirect thought as well as indirect speech (Gedankenreferat, Redereferat), takes up the Lubbock distinction between the scenic and panoramic description and divides what I prefer to call comment up into (...) commentary, analysis and reflection."
Anglo-American and French Studies
- Thale (1967): reportorial, enumerative, selective, atmospheric, metaphoric, hyperbolic.
4- "These terms, however, fail to meet all three conditions by which we judge a list of modes: balance, differentiating power and inclusivity"
- Bain (1967): 1) narration, 2) exposition, 3) argumentation, 4) evaluation. 1&2 are report, 3&4 comment.
- Kinneavy: narrative classification and evaluation... (Petsch 1934 is better than both!)
- Chatman (1971) relates telling/showing to Plato and to Benveniste (not the same!); Benveniste distinguishes mediated/unmediated narrative; this is not clear-cut, and all narrative is mediated. Existents / Events in Benveniste (= objects of description / of report). This leaves out 3& 4. Chatman 1975, problem of mediation: the bare description of physical action is felt to be essentially non-mediated"; for Benveniste this is true of unmixed report. Mixture of narrative modes to suggest mediation - through point of view techniques narrative is felt to be more direct.
- Hernadi (1961): 16 modes (thematic, lyric, and narrative), 4 in narrative: survey, direct quote of speech, interior mologue, substitutionary narration. The list is too complete to apply [?!]
- Todorov, Barthes, etc. : Showing/Telling, Histoire/Discours [Bonheim does not pay attention to Austin - JAGL]
- Genette (1969, 1972):
- Logos (thing said)
vs.
- Lexis (manner of saying):
- Mimesis
vs.
- Diegesis:
- Description
vs.
- Narration:
- Narrative (report) vs.
- Discourse (comment)
7- (On this last, cf. Benveniste, and Weinrich's "erzählte Welt" / "besprochene Welt": "narration as opposed to discussion"). Genette is still based on "the supposition that narrative is not mimetic if it takes a form other than speech (...) The system rests on a distinction between matter and manner which was thrown out of court by the new critics in the thirties and forties."
Structuralism is deductive; our own 4 modes are inductive. They categorize referential communication, but also emotive, phatic elements, etc. (cf. Jakobson's functions of language). They are not static, they are subject to historical change (cf. Lodge's "modes").
Hierarchy of Modes
Speech is the favourite mode today (the "sense of immediacy" is perhaps appreciated)—it is the peak of the "exit author" thesis. Endings with speech are twice as likely in the 20th century than in the 19th c. Vs. the 19th c., in which description was predominant.
9- Before romanticism, scene description played practically no role in the novel, they were "crowded chronicles of event." From the 16th c., moralizing comment had been foremost, with theories of the instructive end of fiction. Description: painter; Speech: dramatist; Comment: preacher. Report, then, is the essential mode of fiction—but of history too.
10- "Report is the staple mode of narration in all ages." Openings, endings, etc., with a special mode, are tributes to the taste of the time. "The modal façade is a device of narrative which has no apparent organic relationship to the structure behind it." Obeisance to the reader. Often it is a false front, for immediate effect (esp. in the short story, rare in the novel).
The Articulation of Modes
10 - This is also subject to change. "In earlier narrative it is often easy to say to what mode a particular sentence belongs. In our age this is not so." (Leisi). "It has become increasingly difficult to separate the narrator's perception and comments from those of the characters"; cf. Dolezel's neutralization of oppositions. In Plato, 2 modes, easier to handle; compare Mas'ud Zavarzadeh's phenomenalistic or physical vs. noumenalistic & metaphysical realms. Bonheim's reflection on these: Description (space) and Action (time, report, speech) / vs. Comment (reflection). "Plot" or "Character" are not on the same hierarchical level for Bonheim.
|
Time
|
Space
|
Report
|
+
|
+
|
Speech
|
+
|
-
|
Description
|
-
|
+
|
Comment
|
-
|
-
|
In Plato's terms, the first three would be elements of mimesis; comment is ideational.
Elements outside the space/time continuum of the fictional work are not covered (metanarrative elements). Does storytelling develop? Away from metanarrative, away from authorial comment, away from the description of landscape.
13- "We can, at any rate, measure developments and see where the text before us at the moment makes use of the convention in favour of one region or age as opposed to another" (which can be used in a diachronic consideration of texts). Speech, a theatrical element, is absent from the origins of the short story.
Imbedded modes
Speech is easiest to identify, but it is also the mode which most naturally contains elements of the other three. Others are less inclusive but more amalgamative.
Chapter II - Mode Chopping
Bonheim denies the practical applicability of earlier models of analysis. Vs. the use of summaries as a basis for analysis, and also vs. too painful exactitude. This does not cover metanerrative or non-representational segments.
The Definition of Modes
Prevalent modes tend to be introduced sooner. An intuitive feeling is often enough. Definition: with linguistic markers of shift.
Speech
The novel: interest in psychology, internalization (from the 18th c.?).
20- "The development called internalization is not, as is often assumed, a matter of an increasing interest in human psychology. Rather, it is a successive shift of techniques, away from the an-at-arm's-lengh authorial depiction (description) and authorial discussion (comment) to a use of the more dynamic modes, especially the submodes of speech." Also, dramatization.
Linguistic marks of speech, etc. (of direct speech vs. indirect speech). It is important that indirect speech is a selection from what was supposedly said. Direct speech is not central to the genre (the 19th c. did without it).
22- Vs. report? "Probably a work of fiction which contains little report will tend to slip off into another genre"— a sketch or essay.
Report
Time-markers, "especially when preceded by description"; action verbs.
22- "There is often little difference between report and description: the depiction of a person at rest is, as we have seen, considered description, the depiction continued when that person begins to move is report. The element of volition is also important (...)". But Bonheim vs. van Dijk's identifying "action" with "intentional".
23- "Nor is it quite just to call any depiction of an event in time report, in contrast to the supposed timelessness of description"; "Report always has an element of time connected with it, description may or may not."
Chatman's and Todorov's divisions (existents/events & interpretive sentences / functions) are not usable in practical criticism. Genette also there are no borderline cases between discours and récit. Report best described as featuring special verb tenses and time markers.
Description
The naming of the existence &/or qualities of something at rest or not moving purposefully. Kinds: 2 criteria: Form / objects.
FORMS of description: Expository or fused. Expository is less common now, and appears later in the text.
24- Fused: "worked into passages of other modes at the word or phrase level".
Genette's 'decorative' vs. 'explicative or symbolic' description is a distinction hard to apply, subjective. If all is a system, then with 'ornamental' we are merely rejecting what was once in favour, and we praise new styles. Zavarzadeh: 'noumentalistic' (revealing the general truth of things) vs. 'phenomenalistic' (merely registering phenomena for their own sakes). Also problematical.; danger of confusing what it is with what the author intended it to be (the intentional fallacy).
OBJECTS: Places (old fashioned), persons, things (rare).
25- Also, "the location of an action in time often seems to be part of an expository description"; this is often a bridge to report. Place is often used as a "figure of testimony" (linked to real existence), and time too, with a gradual progression to fiction.
"Mimetic" elements ("mimetic" used in a sense opposite to Plato's) include: Description of place, of time, person, reports of actions, dialogue, interior monologue. The differences in times, places, and persons of descriptions often define the subgenres to which the work belongs, and build up the reader's horizon of expectations. Not thing Bildeinsatz (cf. Achilles' shield): is it decorative? Noumenalistic? Difficult to say.
Difference whether a description is a perception or an apperception (wilful, conscious reflection on perception) - Hemingway vs. James. Swift changes in Joyce's burial chapter. Apperceptive description shades off into commnent. Categories used in analysis are best chosen according to the particular work analyzed. [The opposite strategy also provides relevant information - JAGL]
Comment
Comment is easy to identify at sentence and paragraph level, but problematic at word and phrase level.
30- "We expect this mode to use evaluative modifiers, generalizations not imputed to one of the fictional characters or judgements useing a fairly high level of abstraction." Is metanarrative comment? Best seen as a separate category. Logical connectors at sentence level suggest comment too, and adverbials such as "possibly" and "it might be that": they do not belong to the casual spectator (comparable to the essay, or to asides in drama). Nowadays they are avoided, and can be dispensed with almost altogether.
31- Submodes of sermon apply to comment: Melachton: literal preaching vs. epitreptic preaching (devoted to inducing belief) and paraenetic preaching (devoted to persuading the reader to a course of conduct). Comparable to Dante's literal, allegorical and moral levels (anagogical, beyond life, is rare now).
32- Epitreptic comment may crop up in subtle ways, but out-and-out paraenetic comment is altogether out of fashion.
Borderline Cases
- Thought and perception may fall under the heading of speech. In first person narrative, the narrator's words are excepted (emphasis falls on the narrator's or the characters' consciousness, etc.). [This sounds rather primitive narratologically speaking - JAGL].
- Report is the depiction of action, usually human ("birds singing" is description, not report).
- Comment is more general and abstractive; description must be perceived through the senses.
Problems in mode-chopping
Transformations (after generative grammar: negative, interrogative, etc.) are applied to modes or virtual forms (except in comment):
34- "The virtual form (...) consists of imagined speech, of report conceivable rather than actual, or of imaginary description."
There is an increasing use of virtual modes, linked to internalization, to a focus on the consciousness of the character, and to fabulation [hypotyposis]. [Note that virtual modes are a real but at the level of thought; they still refer to elements of the action - JAGL]
Chapter III - The Modes in Concert
"Transitional probabilities":
38- "Description tends to be followed by report rather than speech, and speech tends to be preceded, and less often, followed by report." Dynamic: Speech is linked to report; Static: commentary linked to description.
Modal deficits
Sometimes authors are characterized by a deficit in some mode or submode (apperception, etc.). Difference: whether the deficit is characteristic of a passage, a work, an author or a period.
Changing Tastes in Narrative Modes
Differences in mediation, narrative pace, selectivity and reader participation have determined the modern shift to dynamic modes.
Modes and Mediation
39- "Telling is less mimetic than showing, a distinction at odds with Plato's, where by definition an undramatized work of literature cannot in the original sense be mimetic." Metanarrative and comment are unmimetic; description is often seen as mediated, with "telltale rhetorical touches", or the way things are named revealing omniscience. Scale of mediation in narrative: metanarrative, comment, description, report, speech.
The Submodes Reviewed: Narrative pace
[See above.] Expository description: adds integral type, which determines the development of action. Comment: pure or integral (i.e. fused with other modes), with opinions attributed to characters...
- Pure comment, expository description, stopped fictional time
- Integral comment, fused description, slow fictional time
- Direct speech, scenic report, fairly fast fictional time
- Indirect speech, panoramic report, fast fictional time
(With 2 submodes each at least).
42- "The illusion of pace is caused by narrative time being telescoped, stretched, or balanced with narrative time, and two of the submodes allow such a balance: direct speech and scenic report" (favoured nowadays). [But there is, rather, an illusion of coincidence - JAGL]
Problems of representation [Too sketchy a section - JAGL]
The Selectivity Principle
The high technical level of the modern novel must seem careless and unselective. Patently selective modes and submodes are less favoured.
Reader participation
Speech and Report are "cool" modes (McLuhan) - they invite participation. But it is usually Comment and Description that are supplied to the reader, not speech and action. Modes require little response in the same mode, but leicit it in the others.
Chapter IV - The Submodes of Speech
2 axes: direct/indirect speech and outer/inner speech: resulting in 4 basic kinds: direct speech, thought, indirect speech, and substitionary thought. But extended: (Bernard Fehr, "Substitutionary Narration and Description: A Chapter in Stylistics", English Studies 20 (1938): 97-107).
50- "Fehr saw that a character's consciousness (...) may take the form of perception as well as thought (...) The third plane beyond speech and thought Fehr called 'substitutionary perception', a concept taken over by Hernadi [1972] and others."
- There is a whole axis between direct and indirect speech (Page).
- Direct or substitutionary thought and direct or substitionary perception are submodes of speech; analogous to comment & description/report in the narrator. [Bonheim distinguishes between direct style, free indirect style, and indirect style, but only two kinds in thought and in perception. P. 55: three types too.]
Distance
[Here Bonheim is a mixture of Booth and Genette - JAGL]. 6 kinds of distance, between author, narrator, character and reader. Submodes of speech are a distance between reader and character.
He distinguishes with Pascal (1977), Cohn, and others, direct from indirect interior monologue. Degrees of distance. Minimal number useful, 3 degrees. (A table here with the equivalents in other critics). Direct speech is the most indirect way of narrator's speech.
Direct speech: special cases or manipulations:
- traslated speech
- coloured by the idiosyncrasies of the speaker
- sometimes meant to be a selection
- filtered through another character's consciousness (& shifts, etc.).
Marks may be more direct in the narrator. Deviant kinds are sometimes called "free direct speech" (e.g. without an inquit, or without graphological indications, or midway towards indirect speech...) In the 18th c., often quotation marks are used for indirect speech as well. Degree of mutation of the speaker's voice, shifts, etc. [sometimes confusing - JAGL]. Scene involves coincidence of times; panorama involves "ruffles".
Submodes of thought. Direct thought, often with quotation marks. Syntactical fragmentation is usual ('atactic' texts), and kinds of signals for narrated monologue (punctuation, inquit, tenses, pronouns, deixis of time & space, phonological, grammatical & lexical idiosyncrasies...) Shifts, too. Conventions become more and more comples as the view is interiorized. Writers break the conventions they themselves set. Breaking of expectations is balanced by the needs of grammar.
Submodes of perception.
69- "It is now over 40 years ago that Fehr saw erlebte Wahrnehmung (narrated perception) as parallel to and separable from erlebte Rede (narrated monologue). He also called it 'vision by proxy'" (but applied to all the senses); 69- "The common denominator is the implicit or explicit recognition that most of the information which a narrator conveys to the reader can also be loaded onto the vehicle of characters, who take over one of the jobs of conveying details of the fictional world." But thought does not equal description and report; it is nonverbal (not equivalent to monologue, but often shady). The first perception indicator is felt to govern any subsequent sequence of the same kind. There are unreliable thinkers and perceivers, as well as unreliable narrators. And there is free direct perception too, not narrated off as indirect, and not coincident with the original.
A review of Submode Markers.
There are degrees in all, but there is a set of markers at all three levels: 1) quotation markis, etc. 2) inquits (& percepts); 3) Tenses; 4) Deixis of space and time; 5) Personal pronouns; 6) Relatives; 7) Phonological markers; 8) Grammatical markers; 9) Lexical markers; 10) Rhetorical markers (e.g. lack of connectors, lexis...). There are thousands of possible combinations. Simplification in theory is necessary; technical terms are necessarily overlapping and reductionist. Intelligent reading is necessary to supplement the system.
5. Inquits
Bound (or tagged) speech vs. free speech.
75- "The inquit acts as a hinge between passages of speech and the and the adjacent narrative modes, and belongs properly to the mode of report." Historical considerations are necessary in the system (use declines). Initial inquits in the Renaissance, today more likely in final position. Or done away with (Joyce's influence). Now, shorter inquits and speeches, and author-specific techniques.
Inquit norms.
They are determinable: length of speech (No medial inquits in short speeches; not final inquits in long speeches). Inquits are loaded with descripion, etc. (esp. initial). Shifting from formal to colloquial style (usually the reverse); word order (now inversion pronoun/verb unusual). With nouns, often inversion if modified; inversion is more usual with 'say' (Jespersen); a matter of focus. Topicalization of inquit elements is linked to the position of inquit. 'Crypto-inquits' such as description of character, or 'stage directions'. Now there are less multiple-part inquits. Tense: sometimes contrasts are used; present participle, etc.
Scale of mediacy.
Speech - Report - Description - Comment - Metanarrative. Often, the middle forms are buffers between the first and the last. If zero-stage (no inquit), then possible markers are the change of addressee, the use of pronouns, registers or viewpoints of speaker, crypto-inquits... Redundancy is already present in old writers: now more economical.
The inquit and stylistic norms.
This follows laws of fashion, not absolute. Multiform breaking is now possible. Rules are not the same as norms or conventions. "The inquit conventions are not 'mere conventions' the function of which is obscure, but a set of interrelated norms subject at least in part to authorial manipulation and, therefore, to critical analysis."
6. Short Story Beginnings
Can we speak of closed and open endings and beginnings? Beginnings can have more or less exposition. Openings in the short story are more important than in the novel. And there is less delayed exposition.
Openings with comment are becoming rarer and rarer. With description, frequent, but no longer massive description. Of interiors and city-scapes: the relevance is less clear, it creates atmosphere but not describe the setting of the story. There is a wish to reduce expositional mediacy. Comment and description are not inflexibly linked to exposition; report and speech are not necessarily medias in res. Circa 40% of stories begin with report, other modes the rest. A variety of means to suggest habitual action in exposition. Panoramic report is more expositional. The five W's in exposition (when, where, why, what, why) create an unmodern impression. Circa 11% begin with speech, often a façade, with exposition following.
Modes imbedded in speech. Speech has more functions: all of Jakobson's functions. (Report, description and comment are subfunctions of the referential function). Functions of speech are usually hard to define—see also Searle's speech acts:
114- "If we want to categorize all forms of speech and its submodes, then, we probably need not only the modes and language function of Jakobson's model but speech act theory as well."
Only in the mode of speech is the whole gamut found. [Sometimes] speech sends signs to reader that he will have no exposition.
7. How Stories End: The Static Modes
Endings are more significant than beginnings (cf. Poe on Hawthorne)—and more varied. Open and closed endings. Action may simply stop; vs. signals used to announce end: repetition of previous elements, a symbolic event... When action stops conflicts are left unresolved. Now endings are more or less closed. The shift to present and metanarrative are the most closed ones (cf. Labov's coda). Now this is a taboo in third person narrative. Comment is also dispensed with, or distanced with irony and other techniques. But "we have come to expect a story to speak for itself" —> hence, dynamic modes are favoured. Comment becomes 'interiorized' or of particular, not universal, application; or it may come from a chorus character. If interiorized it is more linked to the main fabric. Description is also a closed ending. More closed if it is long, of place, modally pure, and unfunctional; it acts as a frame. Death as a subject is associated to closed endings; rhetorically heightened language, too.
8. How Stories End: The Dynamic Modes
135- "The dynamic modes, report and speech, by contrast, are conducive to an open effect" unless they are mixed with the others. Open endings are parallel to in medias res beginning. Openings are modified by by a series of elements, and endings with death (rarer now) have less overt pathos, or are announced, or the character simply departs. (Cf. epanalepsis).
Endings which repeat beginnings. They often repeat titles. This is conventional, and reduces openness. Sometimes we have 'Janus-headed' endings, both looking backward and forward. Or with polyptoton: x.... x'..... This is most often found in endings with report. There is rarely a purity in open or closed endings.
Mitigation of the open effect. With telling, with panoramic report (tableau, ultimate form). Preparation of the end, with comment, etc. are mitigations of the open effect.
Considerations of style. Diction, syntax & rhetoric:
Diction: It is more casual in the 20th than in the 19th c. As a rule, there is a shift upwards at the end.
Syntax (normal, inverted, elaborated, or fragmentary). Deviation at the end from any one of those to the others.
154- "And-" or "But-" clauses are common at the end for a solemn close. Much used in oral style, but not in literature when ending in speech. Play with conjunctions creates a sense of an ending.
156- Irony also occurs rather at the end (—in the endings of novels there is no sudden irony); this is mostly in speech and a modern development.
157- Report: Paradoxically, open ending: both a modern ideal and goes against the tightly-knit structure of the work. All endings are seen as contributing to the whole. "So the critic who claims to have found a genuinely open endingis in effect confessing his inability to interpret it". Openness is a function of the reader's response, not only a property of the text.
Unreliability of narrators or informers. Endings with speech in about 1/3 of stories; short questions with 'and' or 'but', as 'afterthought', are increasing in history.
161- In the 18th and 19th centuries, "By and large the narrative was synthetic rather than analytic: it developed from a set of premises in logical and chronological order, rather than from a puzzle to be solved."
Virtual modes: substitutionary speech and perception, etc., to 'surfiction' or 'fabulation'. This would break our modes, but it is not yet seen in short stories in the 1970s.
Conclusion: The Novel and the Short Story
Bonheim favours categorization and theory as a proactive invitation to the avant-garde and creative writing. Differences: vs. Poe's and Mathews' short story theory; vs. the principle of a single setting and explosive principle to define the short story. There are no binding laws: "a short story is a complete narrative, normally too short to be published by itslef" (Aristotelian definition of short-story plot); closer to tragedy or drama than to longer narrative. Differences in closing: short sentences, speech and descriptive sentences appear mostly in the short story. But epanalepsis too, unexpectedly. No ironic endings in novels. Analysis of genre terms for novels—there is here an overlapping with the short story, both in genres and in techniques.
—oOo—
Continuará....
