Invention and Transmission : Seymour Chatman's Narrative

Title Theory

Author(s) Takeda, Masafumi

Citation 待兼山論叢. 文学篇. 31 P.1-P.15

Issue Date 1997

Text Version publisher

URL http://hdl.handle.net/11094/47907

DOI

Rights

Osaka University
 1

Invention and Transmission:

Seymour Chatman's Narrative Theory

Masafumi TAKEDA

Seymour Chatman, 1) in his Story and Discourse: Narrative Struc-

ture in Fiction and Film (1978), integrated continental narrative the-
ories of Roland Barthes, Tzvetan Todorov, Gerard Genette, and Anglo-
American critical tradition of Henry James, Percy Lubbock, and
Wayne Booth. The French word "N arratologie" was transplanted to
English soil as "Narratology", the theory of narrative. Chatman's
primary interest focuses on the question: "What is narrative per se?
What properties must a text have to be called a narrative, and what
properties disqualify it?" From the fact that the same story can take
various forms (novel, play, film, ballet, comic strips, and so on), narra-
tive is assumed to have a deep structure which is quite independent of
its medium. The deep structure consists of story elements, which
different modes of representation convey to the audience. His
approach, as he openly admits, is "dualist" and "structurist." This leads
to his examination of the surface and deep structures of narrative, his
dichotomy of story and discourse, that is, the what and the how - the
story content (actions, characters, settings, themes) and the means by
which the story is presented. His work was a systematic attempt to
probe into the narrative structures of story and discourse.
In Coming to Terms: The Rhetoric of Narrative in Fiction and
Film (1990), Chatman refines his narrative theory with notable develop-
ments as well as some modifications of his former work. In response
to the criticism against Story and Discourse that his accounts of film
are too scarce, he offers extended investigations of cinematic descrip-
tion, narrator, and point of view, and allots one whole chapter to the
analysis of John Fowles' film version of the famous novel, The French
Lieutenant's Woman in detail. The analysis of different discourses -
2

novel and film - representing the same story helps to clarify the
distinctive features of the two modes of discourse. My concern here,
however, is with his newly developed analysis of the narrative transmis-
sion on the discourse level in verbal fiction, especially novels. In
addition to the redefinition of important narratological terms, Chatman
describes more minutely the positions and functions of narrative com-
ponents such as implied author, narrator, character, and implied reader,
in the context which he calls "invention" and "transmission" In this
paper I will review his theories of Narrative-text and the narrative
components on the discourse level, and the problems with regard to the
interpretation of narrative on the basis of his transmission model.

Definition of Narrative

For a systematic study of narrative, what is needed first is the

definition of narrative. In Story and Discourse, Chatman analyzes the
structure of narrative in his strictly binary-oppositional way: Narrative

Actions
Events {
Happenings Form
of
Characters Content
Story Existents {
(Content) Settings

People, things, etc., as pre-} Substance

processed by the author's = of
cultural codes Content
Narrative
Form
Structure of
narrative transmission }= of
Expression
Discourse

{ ~:~:atic
(Expression)
} Substance
Manifestation Balletic . . = of
Pantorrurruc Expression
etc.

(ND, 26)2l
 Invention and Transmission 3

is divided into Story (content) and Discourse (expression). And Story

and Discourse are respectively divided into Substance and Form.
On the other hand, in Coming to Terms, he considers narrative
from an external viewpoint, that is, in terms of the relations of
Narrative to other kinds of text-type, such as Description and Argu-
ment.

Text

Narrative Argument Description Other

~
Diegetic
Mimetic

Nov~tc. Pla~oon, etc.

(CT, 115)
Description is a text that renders the properties of things either visible
or imaginable. Argument is a text that attempts to persuade an
audience of the validity of some proposition. The property of Narra-
tive is "its doubly temporal logic" that entails the external time (the
duration of the presentation of the novel, film, play) and the internal
time (the duration of the sequence of events that constitute the plot).
(CT, 9) Chatman's point is that while literary works are generic
subclasses of the Narrative text-type, fictional narrators utilize the
other two as well in narrative transmission. These three kinds of
discourse "operate at each other's work" in fiction as well as in every
situation where writers or speakers are involved in the acts of descrip-
tion and argument.
The narrator uses Argument to give commentaries on actions or
characters and his own discoursive manner, sometimes as a form of
maxim and generalization. On the other hand, Description, which
conveys the qualities attributed to a story-object, has more subtle
differences from Narrative. Description is rendered (1) as assertions
("Simon is simple."); (2) as nonassertive mentions or inclusions ("Simple
Simon .... "); (3) as elliptical implications ("A passerby asked Simon for a
4

shilling, and Simon gave it to him. The passerby laughed and ran off.")
In the last example, the reader infers his simplicity by interpreting the
two sentences, according to codes which he believes to operate in the
context (a capitalist code in this case, and there are other possibilities).
And we should note that Description can render action, but fundamen-
tally in a different way from that of Narrative: the description of action
functions only as part of the narrative settings; on the other hand, the
narration of actions must contain the dual-time logic and present an
action that leads to the event chain of plot, which is of narrative
significance. Thus, the examination of each text-type of the sentences
makes it clear that verbal structure can direct us to the narrative
structure.

Discourse Elements

Next, I would like to review the discoursive components of

narrative. In Story and Discourse, Chatman offers the following
much-cited diagram:
Narrative text

Real Imthplied-+(Narrator)-+(Narratee)-+lmplid.ed Real

---~
author ---~ au or rea er reader

(SD, 151)

The Real Author

In Chatman's theory, real life authors and readers are quite alien
to textual communication.. One of the advantages of the notion of the
implied author is that it enables us not to commit "intentional fallacy."
If the same writer's several works are quite different or even contradic-
tory_ in themes, itis_meaningless in a narratological sense_ thaLthe
reader tries to attribute the thematic shifts to certain transitions of his
personal views or moods, which we canrtot exactly understand from
reading his biography or from knowing him personally.
Chatman explains the stance of a narratologist:
Just as linguistics argues for a logical model, not a behavioral account of
 Invention and Transmission 5
actual speech performance, narratology offers a theory which assumes
the task of defining its subject (all and only narratives in the universe of
texts) on a logical model, with no reference to the contingent life histories
of those who make or partake of stories. (N, 261)
Narratological text theory presupposes that the source of the genera-
tion of meaning is the text itself, and therefore, "Real authorial behav-
ior is a subject for literary biography, not text theory." (CT, 80)

The Implied Author and the Narrator

When Wayne Booth used the term "the implied author," he meant
that it was the real author's "second self," "an implied version of
himself," which the reader constructs from the text. 3) The emphasis
was placed upon the value-laden picture of the author which was proper
to a particular text only. When narrative theorists utilized the notion,
it was incorporated into narrative communication model. They em-
phasized the importance of the notion of "implied," that is, the reader's
act of inferring the author's image, norms, or intent.
For Chatman, narrative is "an invention, by an implied author, of
events and characters and objects (the story) and of a modus (the
discourse) by which these are communicated." The implied author is,
so to speak, "a guiding intelligence" and the narrator is "a means or
instrument." The implied author assigns to the narrative agent the t(;lsk
of articulating story elements, of actually offering them to some
projected or inscribed audience (the narratee). While the narrator is
only the transmitter of the story, the implied author is responsible for
"its whole design." But "there is some contentual disparity between
text's intent and narrator's intention." (Here, "Intent" means a work's
"whole" or "overall" meaning, including its connotations, implications,
unspoken messages.) (CT, 74) As he puts it, "there is a good reason in
theory to keep invention and transmission separate as text principles."
The theoretical distinction is essential because the two terms explain
different levels and sources of information. The source of a narrative
text's whole structure of meaning-·- not only of its assertion and
denotation but also of its implication, connotation, and ideological
6

nexus - is the implied author. (CT, 75-6)

In verbal fiction, the act of the narrator is a surface structure, and
a deep structure is the story components. The representation by the
narrator of the story generates meaning, but the meaning may be
supported, questioned, or ironized by the underlying implied author's
distinctive meaning. It is the central point of Chatman's narratological
theory that the two text principles of invention and transmission func-
tion separately in the communication or actualization of narrative text.

The Implied Reader and the N arratee

The implied reader is, like the implied author, not a real reader,
but the reader who is presupposed by the narrative text. It is construct-
ed by the implied author's values or norms; put in another way, it is the
reader who can understand the implied author's intent. On the other
hand, the narratee is a counterpart of the narrator, one who is addres-
sed by the narrator. When one of two characters speaks to the other,
we can easily identify a narratee. Though, in many cases, the narrator,
standing outside the story world, seems to addresses directly to the
reader, theoretically the narratee is inscribed in the narrative text and
must be distinguished from the implied reader.
The implied reader is required not to take the narrator's meanings
literally, but to seek deeper meanings which the text implies. Accord-
ing to Chatman, the implied reader's task is to infer and reconstruct the
implied author's intent from the text, developing assumptions that lead
to a satisfying interpretation.
Then, the question is, how can we discern the implied author's
intent from the narrator's intention? The narrator does not always
reflect the values of the implied author, especially in unreliable narra-
tion. By definition, the implied author does not "speak," does not
deliver direct messages to the reader. "The implied author only implies
messages, and we understand those messages only by inferring them
from the total fiction - not only from what the narrator says, but
from what happens, what the characters are like, what they say about
each other, what the setting and atmosphere suggest, and so on." (RNF,
 Invention and Transmission 7

242) This leads to the problems involved in the interpretation of

narrative texts, which will be discussed later. Though the structuralist
reading is ultimately to find meanings from the relationship of narra-
tive components, Chatman's analysis of narrative structure seems to
center on the production of meaning among discoursive agents.
The recognition that the represented words and deeds of the
characters get transformed and may well be distorted by the narrator
always gives us clues to the discovery of new meaning. In the actual
reading of the text, we ask ourselves: Why does the narrator choose
this particular word? How does the narrator direct the reader's
attention toward or away from this? Do we gain equal access to the
minds of each character? Why does the narrator use a flashback?
Taking into the consideration the basic distinction of story and dis-
course, and the arrangements of the story elements in the narration, we
can fix our thought on rhetoric from the "choices" of the narrator,
under the control of the implied author. 4)

Slant and Filter

Narratology originally began with the study of narrative

content(such as Propp's study of folktales), but after Barthes' Introduc-
tion to the Structural Analysis of Narratives (1966), the interest was
directed toward the form of narrative, the way how the story is
communicated. There are two mainstreams of narratology: the
thematic study of narrative content and the formal study of narrative
mode of representation. According to Martin's further classification,
modern theories of narrative fall into three groups, which treat narra-
tive in terms of (a) plot (sequence of events), (b) point of view (a
discourse produced by a narrator), and (c) reading (a verbal artifact that
is organized and endowed with meaning by its reader). 5) According to
his classification, one would say that Chatman belongs to the second
party.
As discussed earlier, Chatman's distinction between the implied
author and the narrator, in the most basic sense, serves to separate the
textual denotation (what the narrator says) from the connotation (what
8

the implied author means). This point is more clearly explained by his
refined discussion of narrative perspective.
The terms "point of view" and even "focalization" are insufficient
to cover both the narrator's and characters' different mental acts.
Originally the term "point of view" has two senses: "a point from which
things are viewed" and "a mental position or viewpoint." It can be used
both literally and figuratively. To avoid confusion of the narrator's and
characters' mental behaviors, stances, attitudes, and interests, Chatman
proposes the separate names for the two. The attitudes and mental
nuances of the narrator who is in the discourse world is "slant"; the
mental activity characters experience in the story world is "filter." (CT,
143)
"Slant," Chatman writes, "catches the nuance of the choice made
by the implied author." The implied author, explicitly or implicitly,
manipulates the narrator's psychological, sociological, or ideological
attitudes to "illuminate" or "keep obscure" particular aspects of the
story world. "Filter" belongs only to the story world, indicating the
various mental attitudes of a character toward the story-objects. Here
Chatman's theory of narrative invention makes quite an important
distinction. When a character plays a role of a narrator, what we get
is his slant, not his filter, because his narration is performed on the
discourse level. Theoretically, what the character-narrator presents
(what he tells, shows, or reports) is nothing but his reflection of the
original events that he experienced or is experiencing. When he, as a
narrator, presents the original events, there is always a possibility that
they are subjected to some transformation or filtration, (though this is
also, in the final analysis, manipulated by the implied author).
Narrative text can include two kinds of "untrustworthiness."
First, the narrator's account of the events or characters seems at odds
with what the text implies to be fact. Second, a character's speeches
and thoughts about the story events or other characters seem at odds
with what the narrator is presenting on the discourse level. Chatman
names the former "unreliable narration" and the latter "fallible filtra-
tion." (CT, 149) The distortion of story information by the narrator is
 Invention and Transmission 9

attributed to various traits of his personality - his cupidity, cretinism,

gullibility, innocence, inconscience, or a lack of information, and so on.
(Remember that these are also assigned to the narrator by the implied
author, the inventor. The limitation on the narrator's capacity to
render the details of the story world depends not on "knowledge" but on
how much the implied author has delegated to him to present. (CT, 121))
"Fallible" is used to refer to the character's inaccurate, misled, or
self-serving perception of events, situation, and other characters.
Never conscious of being reported by the narrator and having no direct
access to the transmission of the story, the character cannot be charged
with unreliable narration, and the term "fallible," which means "liable
to mistake or to error," is appropriate for the character.
Unreliable narration and fallible filtration can be conveniently
explained in the transmission diagram.

Implied Arbor
I
Narrator~ Story r Narratee Imflied Reader
I
L---~-------------~
(CT, 151)
The fallibility of a character is shown by the narrator explicitly or
implicitly. An reliable narrator could explain and comment on it
clearly, but in many cases, the narrator covertly presents it seemingly
in an objective manner. Between the narrator and the narratee, the
narrator points out a filtered view of a character, to make the narratee
enjoy an irony at the expense of the character. On the other hand, the
narrator's unreliability is always implicit, and it must be inferred by the
implied reader with the implied author's secret message. If the reader
perceives two conflicting messages in what a narrator says (one is an
ostensible meaning of the narrator; ihe_otherJs. an implicit meaning
which he is never conscious of), the narrator becomes the target of the
irony between the implied author and the implied reader.
What Chatman clarifies by his renewed theory on perspective is
the more explicit distinction between the narrator's act on the discourse
level and the character's act on the story level, and the power of
10

narrative transmission to generate meanings on different levels through

its components.

Problems in Chatman's Theory

As part of theoretical model, the notions of the implied author and

the implied reader are necessarily abstract ones. Genette, who com-
pletely rejects Wolfgang Iser's notion of the implied reader, also feels
the necessity of correcting the image of the implied author. Though he
agrees that such a construct as the implied author is necessary to
reading, he complains, "if one wants to establ_ish this idea of the author
as a 'narrative agent,' I don't go along, maintaining always that agents
should not be multiplied unnecessarily."6) Taken in terms of production
and reception, the implied reader is the idea, in the real author's head,
of a possible reader and the implied author is the idea, in the reader's
head, of a real author. For Genette the implied author is almost an
equivalent for the real author, or at least the "induced" author and the
implied reader is only the possible or "potential" reader. Thus he
insists that the implied author and the implied reader, as irrelevant
agents, should be excluded from the communication model. But as he
asserts that his study is limited to the narrative discourse and the
narration and he has no concerns for its objects (the story elements, and
norms of text, the deficiency of which in his study led to Booth's
criticism), what one includes and excludes in the model depends on
one's own purpose for discussion.
Shlomith Rimmon-Kenan raises the objection to the notion of the
implied author, in that if it is only a construct which has no voice, no
direct means of communicating, it is a contradiction to assign to it
some kind of a role as an addresser in a communication situation. She,
instead, proposes to de-personify it and understand it as "a set of
implicit norms." 7) In Coming to Terms, Chatman claims no insistence
on referring to it as if it were a human agent: "That inventor is no
person, no substance, no object: it is, rather, the patterns in the text
which the reader negotiates. He is willing to substitute "text principle,"
"text design" or "text intent" for the "implied author." (CT, 86-7)
 Invention and Transmission 11

Another criticism of Rimmon-Kenan's is directed to the existence

of the narrator and the narratee. In his model in Fiction and Discourse,
Chatman put the narrator and the narratee in brackets as "optional," on
the grounds that in the cases where the narrator seems to be effaced or
disappear, as in Hemingway's "The Killers" and the pure dialogue short
story, such narratives are "non-narrated." Chatman revises his view on
the narrator, saying that every narrative is by definition narrated
(narratively presented), and that the story is "shown" by a silent,
extradiegetic narrator. The narrator's task consists of not only telling
but also showing. "Any narrator, whether authorial, camera-eye, or
dramatized, is a tool of the invention." (CT, 85)
Finally, Michael J. Toolan also offers a negative view on the
status of the implied author.
The implied author is a real position in narrative processing, a receptor's
construct, but it is not a real role in narrative transmission. It is a
projection back from the decoding side, not a real projecting stage on the
encoding side.8)
What is common to the criticisms of the three critics is the problem
that the implied author is never a narrative transmitter. As Chatman
admits that the narrator is the only voice of narrative discourse, and
that the implied author is what the implied reader infers and recon-
structs, they are right in saying that the implied author can have no
position in the transmission process. But in Chatman's theory, narra-
tive entails two separate text principles, invention and transmission.
Because as the inventor, the implied author is the source of a narrative
text's whole structure of meaning, and because Chatman excludes the
real author for his resolute reasons, there needs absolutely to be an
agent before (behind) the narrator. With regard to transmission, though
the implied author only implies messages, or though he emerges only in
the implied reader's act of reconstruction, he manipulates from behind
everything the narrator presents. What I mean is that there are
communications between the narrator and the implied reader, and
between the implied author and the implied reader. The direction of
arrows is trifling; the most important thing is that Chatman's model
12

indicates clearly that the implied author functions as both the inventor
of narrative and as the medium of the implied reader's act of com-
municating in narrative text.

In terpreta ti on

Finally I consider the problem of interpretation of narrative texts.

In the act of interpretation, we regard the implied author as an author-
ity. If we do not posit such a construct, we have to search for an
authority either in the narrator or the characters of the text, or in
something outside the text, in a wide range of social, ideological,
historical sourses. In the recent critical situations, the interaction
between the text and the critic constructs a number of critical
narratives. 9l The exclusion of the real author and the real reader is
bound to limit the possibilities of interpretation. But it is no doubt true
that structuralist reading will always be useful for close reading.
But uncertainty lies in the text-centered criticism. If we hope for
any certainty in the text, narrative uncertainties emerge not only in
narration but also in the principle we attempt to infer behind the
narrator. It could be said that "it is precisely the unreliability of the
implied author that has come to typify post-modernist narrative." 10)
Poststructuralist theorists have pointed out a crucial contradic-
tion immanent in narratology, called "double logic of narrative." 11 ) One
principle emphasizes the priority of story over discourse, insisting on
events as the origin of meaning; the other stresses the priority of
discourse over story, insisting on events as only the product of dis-
course. Because the two contradictory principles can never be synthes-
ized, narratology will always be deficient. Jonathan Culler calls this
narrative problem "a relation of dependency," which contains in narra-
tive "a self-deconstructive force." 12 l When one discusses story and
discourse in narrative, one must first determine which is the given and
which is the product, but either choice will miss the curious complexity
and the impact of narrative. This concerns a matter of hermeneutic
practices, but as Culler suggests, apart from the possibility of synthesis,
the reader should shift from story to discourse, and from discourse to
 Invention and Transmission 13
story.
Chatman seems to be inclined to stress the narrative discourse,
and more interested in the production of meaning through the narrative
transmission. Structuralist theories have been criticized for being
static and unable to capture narrative dynamics (in addition to igno-
rance of history). But through the shifts of perspective, from discourse
to story, and from story to discourse, narratology can grasp narrative
dynamics, such as tension, suspense, and surprise. And Chatman's
theory, though in fact never too theoretical, makes enough room for the
thematic interpretation, because the implied author can function as a
text's whole structure of meaning.
Admitting that the scope of narratology is a limited one, Mieke
Bal states that she has used her narratological theory for both aesthetic
and ideological (feminist, in her case) criticism. 13) Chatman's narrative
theory also can be utilized in reading narrative texts in general, as "an
indispensable tool."
My defense [of the implied author] is strictly pragmatic, not ontological:
the question is not whether the implied author exists but what we get from
positing such a concept. (CT, 75)

Notes
1) Seymour Chatman, Professor Emeritus of Rhetoric, Professor in the
Graduate School at University of California, Berklely, specializes in
narrative structure and style in literature and film. His recent study
is targeting the analysis of narrative devices in film and other
popular media. He is the author of The Later Style of Henry James
(Greenwood Pub Group, 1972), and Antonioni: Or, the Surface of the
World (University of California Press, 1985).
2) Page numbers of quotations from Chatman's works are indicated in
the text with the following abbreviations.
SD: Story and Discourse: Narrative Structure in Fiction and Film
(Ithaca: Cornell University Press, 1978)
CT: Coming to Terms: The Rhetoric of Narrative in Fiction and
Film (Ithaca: Cornell University Press, 1990)
14
RNF: Reading Narrative Fiction (New York: Macmillan Publish-
ing Company, 1993)
N: "What Novels Can Do That Films Can't (and Vice Versa)" and
"Reply to Barbara Herrnstein Smith," in On Narrative, ed. W.
J. T. Mitchell (Chicago: The University of Chicago Press,
1980), pp. 117-36, 258-65.
3) Wayne Booth, The Rhetoric of Fiction. 2d ed. (Chicago: University
of Chicago Press, 1983), p. 71.
4) Patrick O'Neill, Fictions of Discourse: Reading Narrative Theory
(Toronto: University of Toronto Press, 1996). Treating the implied
author as "a director" and the narrator as "a performer," he briefly
summarizes the elements of narration on the standpoint that narra-
tion is the process of transforming story into text, by way of the
implied author's acts of arrangement:
(a) chronologization (the agreement of time, transforming
action into plot.
(b) localizaiton (the arrangement of space transforming space
into setting)
(c) characterization (the arrangement of personality traits, tran-
sforming actors into characters)
(d) focalization (the arrangement of narrative perspective)
(e) verbalization (the arrangement of words on page, making all
of the implied author's arrangements known to the reader-
and duly received by the reader as the 'voice' of the narrator)
(f) validation (of the narrator's degree of reliability) (p. 68.)
5) Wallace Martin, Recent Theories of Narrative (Ithaca: Cornell
University Press, 1986), p. 82.
6) Gerard Genette, Narrative Discourse Revisited, Trans. Jane E. Lewin
(Ithaca: Cornell University Press, 1988), p. 149.
7) Shlomith Rimmon-Kenan, Narrative Fiction: Contemporary Poetics
(London and New York: Routledge, 1983), pp. 88-89.
8) Michael J. Toolan, Narrative: A Critical Linguistic Introduction
(London and New York: Routledge, 1988), p. 78.
9) Mark Currie, "A Narratological Reading of 'Snowed Up'," in Liter-
 Invention and Transmission 15

ary Theories: A Case Study in Critical Performance, ed. Julian

W olfreys and William Baker (London: Macmillan Press LTD, 1996),
p.59. Currie says: "where the structuralist was concerned with
structure, the poststructuralist is concerned with structuration, or
the ways in which the text is constructed by criticism."
10) O'Neill, p.70.
11) Gerald Prince, "Narratology," in The Johns Hopkins Guide to
Literary Theory & Criticism, ed. Michael Groden and Martin
Kreiswirth (Baltimore: The Johns Hopkins University Press, 1994),
p. 527.
12) Jonathan Culler, The pursuit of Signs: Semiotics, Literature, Decon-
struction (Ithaca: Cornell University Press, 1981), pp. 186-87.
13) Mieke Bal, Narratology: Introduction to the Theory of Narrative,
trans. Christine van Boheemen (Toronto: University of Toronto
Press, 1985), p. x.
(Graduate Student)