Syddansk Universitet

Autofiction and authorial unreliable narration

Hansen, Per Krogh

Published in:
Emerging vectors of narratology

DOI:
10.1515/9783110555158-003

Publication date:
2017

Document version
Peer-review version

Citation for pulished version (APA):

Hansen, P. K. (2017). Autofiction and authorial unreliable narration. I P. K. Hansen, J. Pier, P. Roussin, & W.
Schmid (red.), Emerging vectors of narratology (s. 47-60). Berlin/Boston: De Gruyter. Narratologia:
Contributions to Narrative Theory, Bind. 57, DOI: 10.1515/9783110555158-003

General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal ?
Take down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.

Download date: 12. jul.. 2018

 Per Krogh Hansen
(Kolding)

Autofiction and Authorial Unreliable Narration

In recent years, the concept of “the unreliable narrator” has been among
the most debated within narrative theory. In the wake of a series of
provocative articles from the late 1990s by Ansgar Nünning (Nünning
1997, 1998a, 1998b, 1999), questions have been asked again and again
regarding on what basis we determine whether a narrator is unreliable and
how broad the scope of the concept is. Is the presence of an unreliable
narrator in a given text the result of an author’s intentional decision, or is
narratorial unreliability a historically variable reader response to textual
inconsistencies and/or changing cultural norms? Does the concept belong
exclusively to fiction, or does it make sense to approach factual or “real”
narrators with the same concepts we encounter in fictional narrators?
In this article1 I will address these questions with reference to the
genre of “autofiction” (a genre parallel to or a sub-genre of
autobiography) with special attention to the Norwegian author Karl Ove
Knausgård’s six-volume novel Min kamp (2009–2011). I want, on the one
hand, to discuss how autofiction makes itself vulnerable to narratorial
unreliability due to the complex truth status of the told and, on the other
hand, to use this genre to question the concept of unreliable narrator in
rhetorical criticism as dependent on an intentional act by an (implied)
author. This is not to say, however, that unreliable narration cannot be a
narrative technique deliberately used by an author in the creative act of
novel writing, but that by using the concept exclusively in relation to this
phenomenon, we lose sight of important alternative (but comparable)
versions. Among these is the kind of unreliability we can encounter in
autofiction.

1
I would like to thank Professor James Phelan (Ohio) for his valuable comments on an
earlier draft of this article.
2 Per Krogh Hansen

1. The Real Fiction of the Self

When Serge Doubrovsky coined the term “autofiction” in relation to his

1977 novel Fils, he defined it, rather paradoxically, as “Fiction, of strictly
real events and facts.”2
Doubrovsky’s work was inspiringly provoked by the fact that Philippe
Lejeune, in his influential 1975 study Le pacte autobiographique, failed
to leave room for blending novelistic fictional writing and factual
autobiography. Lejeune defined autobiography as

a retrospective prose narrative produced by a real person concerning his own

existence, focusing on his individual life, in particular on the development of
his personality. (Lejeune 1989 [1975], 4)

Lejeune stated that if a reader was confronted with a text where author,
narrator and protagonist were the same, he or she would not accept
counterfactual events or incidents as a matter of fiction. “False”
information in autobiographical writing, Lejeune claimed, would instead
be related to “the order of lying” (“l’ordre du mesonge”).
Doubrovsky broke this dogma two years after Lejeune’s study when
he published the novel Fils and proclaimed a new genre: autofiction. For
Doubrovsky, it was characteristic of this new genre that it required
homonymy between its author, narrator and character and that it played
on the generic ambiguity of its contradictory pact: on the one hand, the

2
Doubrovsky coined the term on the back cover of the novel, where he wrote:
“Autobiographie ? Non, c’est un privilège réservé aux importants de ce monde, au soir
de leur vie, et dans un beau style. Fiction, d’évènements et de faits strictement réels ;
si l’on veut autofiction, d’avoir confié le langage d’une aventure à l’aventure d’un
langage en liberté, hors sagesse et hors syntaxe du roman, traditionnel ou nouveau.
Rencontres, fils de mots, allitérations, assonances, dissonances, écriture d’avant ou
d’après littérature, concrète, comme on dit musique.” (Doubrovsky 1977)
“Autobiography? No, that is a privilege reserved for the important persons of
this world, in the evening of their lives, and in a beautiful style. Fiction, of strictly real
events and facts, autofiction if you will; to have given the language of an adventure to
the adventure of a language in freedom, without wisdom and outside the syntax of the
novel syntax, traditional or new. Interactions, threads of words, alliterations,
assonances, dissonances, writing before or after literature, concrete, as they say
music.” (translation mine)
Today, the term is included in French dictionaries, and in Canada it is widely
used as a genre conception, printed on the cover of novels.
 Autofiction and Authorial Unreliable Narration 3

work is claimed to be absolutely referential and factual; on the other, it is

claimed to be a novel, that is: fiction.
The result is not only a contradictory pact but what Poul Behrendt
(2006) has labeled a “double contract.” On the one hand, the text is
subject to the rules governing our social and cultural interactions, that is,
where people have real names and responsibilities, where (at least in our
part of the world) there is freedom of speech, but also laws protecting us
from defamation, etc. On the other hand, the text also relates to the
communicative system of fiction, where reader and author have agreed
that exceptions exist. Here, the distinction between lying and truth is
disregarded: ideas and beliefs can be expressed, even if they are
counterfactual, as contributions to the ongoing negotiation of the
constitution of our culture and “reality.”
Here is not the time and place to follow up on the development of the
concept suggested by Doubrovsky in detail, since it has been scrutinized
and developed by several since then—most importantly perhaps by
Gérard Genette in his Fiction and Diction (1993 [1991]). Genette
suggested that all cases in which an author of fiction includes his own
person (or a character with the same name as the author) in his fictional
story should be considered autofiction. If so, works like Cervantes’ Don
Quixote (1605–1615) and Dante’s The Divine Comedy (1555) would be
included in the category. But as it has been noted, Genette thereby
disregarded the second part of Doubrovsky’s characterization, namely
that the work has to play on the generic ambiguity that comes out of the
claim of absolute referentiality and absolute fictionality at one and the
same time. Illustrated with another example: when in City of Glass
(1985), the first part of Paul Auster’s New York trilogy, a character with
the name “Paul Auster” appears and is furthermore presented as an author
living in New York in an environment that reminds us of the real Paul
Auster’s surroundings, it is less an autofictional than a metafictional
strategy that is being implemented. The intention is not to tell the story of
Paul Auster in fictional terms, but rather to enter a mode comparable to
romantic irony.
What autofiction does is quite radical in the sense that instead of
demarcating fiction from reality it blurs the border. This can be
accomplished by promoting a picture of the authorial self which confirms,
negates, transforms or plays with the public understanding of this self.
Such is the case of Bret Easton Ellis’s Lunar Park (2005), where the
author mixes known public events and incidents from his real life with
counterfactual and even paranormal events and incidents. The blur
4 Per Krogh Hansen

between fiction and reality can also result from using the novel form and
the narrative techniques related to this form to investigate and depict the
history of the author’s own self. This was what J. M. Coetzee did in the
three volumes of autofictional memoirs—Boyhood (1997), Youth (2002)
and Summertime (2009)—when he chose to let the first two volumes be
told heterodiegetically in third-person narration and the third to take the
form of a fictitious biographer’s interviews with five factual people from
Coetzee’s past. Coetzee himself is claimed to be dead in the novel, and
his voice only appears through a number of third-person fragments of the
kind the reader already knows from the first two volumes. But even
though no one would question that the trilogy most certainly is written in
fictional terms and therefore cannot be understood literally, the books are
also understood to be about Coetzee’s own life. In this perspective,
autofiction pushes what is a general paradoxical characteristic of
practically all storytelling, namely that stories tell the (or some) truth,
even though what they are telling might not have happened.
It is therefore also evident that we cannot always rely on the factuality
of the story being told by the author. But can we approach this authorial
unreliability in terms of unreliable narration?

2. Factual Unreliable Narration?

It has been claimed that only in fictional narrative can we have true cases
of unreliable narration. The argument goes that narrative unreliability
depends on, if not difference, then at least on distance between narrator
and authorial agent. In her essay on discordant narration, Dorrit Cohn
claims

that the diagnosis of ‘discordance’ can apply only to a fictional narrative, not
to the kind of storytelling (oral or written) that presumes to refer to real facts:
though we often apply the term ‘unreliable’ to voices we regard as wrong-
headed in non-fictional works (historical, journalistic, biographical, or
autobiographical), the narrator of such works is the author, the author is the
narrator, so that we cannot attribute to them a significance that differs from
the one they explicitly proclaim. (Cohn 2000, 307)

This assumption was challenged by James Phelan in the chapter from

Living to Tell About It on Frank McCourt’s memoir novel Angela’s
Ashes: A Memoir (1996).
Phelan demonstrates how McCourt has “built his narrative on the
foundation of unreliability and the virtual absence of reflection in his
 Autofiction and Authorial Unreliable Narration 5

authorial voice.” McCourt’s “trick” is, as Phelan writes, to conceive

“memoir,” not as an art of direct telling from author to audience, but as an

art of indirection. Rather than speaking in his own voice at the time of the
telling, McCourt uses the historical present and speaks in the voice of his
former self […] at the time of the action. (Phelan 2005, 67)

What makes the narratorial unreliability possible in McCourt’s case is that

he uses what Phelan recognizes as a “nonstandard technique” for his
autobiographical purpose when he “re-invents” his childhood-persona’s
perspective on the incidents. This technique is, on the other hand, a
standard novelistic, fictionalizing technique: McCourt-the-Author has
given the voice and the perspective to a less experienced narrator, Frankie
(McCourt-the-boy-character), and even lets him tell from the moment of
the action that is in the first person and in the present tense. Hereby the
difference or distance claimed by Cohn as a necessity for the formation of
an unreliable narrator is reinstated.
We might therefore also conclude that the extended use of this
fictionalizing technique relates McCourt’s memoir-novel to the genre of
autofiction more than it does to autobiography proper. Angela’s Ashes is
“[f]iction, of strictly real events and facts,” and it suggests homonymy
between author, narrator and character, although this homonymy cannot
be considered absolute: it is nominal and personal, but not temporal.
Furthermore, it plays on the generic ambiguity of what Doubrovsky called
autofiction’s “contradictory pact”: on the one hand, the work claims to be
absolutely referential and factual; on the other, it makes extended use of a
standard novelistic technique and therefore relates to fiction, too, at least
in Richard Walsh’s understanding of fictionality as a rhetorical rather than
an ontological quality. Fictionality should not, in this perspective, “be
equated simply with ‘fiction,’ as a category or genre of narrative: it is a
communicative strategy, and as such it is apparent on some scale within
many nonfictional narratives” (Walsh 2007, 7).
But what would the result be if one’s attention were directed toward a
work that is not composed with the same obvious distance between the
authorial I, the narratorial I and the character I? If we were to follow
Phelan’s line of argument, where unreliable narration is defined as
narration “in which the narrator’s reporting, reading (or interpreting),
and/or regarding (or evaluating) are not in accord with the implied
author’s,” (Phelan 2005, 219) then it is quite unlikely that the narrator
would be unreliable in such a case, since the necessary distance between
6 Per Krogh Hansen

the subject positions is eliminated from the very outset. The question,
then, is whether Phelan’s exclusion would be correct—or whether it is the
result of too narrow a conception of unreliable narration.

3. A Norwegian’s Struggle with his Past and Present

That it is the latter which is the case I will try to demonstrate by looking
at an example of autofiction which makes us consider the reliability of the
narrator without seeking recourse to the concept of the implied author.
The example is the Norwegian author Karl Ove Knausgård’s six-volume
autofictional novel Min kamp, meaning My Struggle.3
My Struggle is among the greatest literary sensations in Scandinavia in
decades. Not only because it has been a huge sales success, but also
because the publication of the six volumes has been accompanied by a
heated debate about the use of autobiographical elements in fiction, and
vice versa. In the six books, Knausgård gives a detailed description of his
life from the day he was born until the moment he types the final sentence
of the manuscript of volume 6, taking into account the reception of the
first volumes of the project and the effect it has had on himself and his
relationships. Knausgård tells the story in the first person and from the
position of the writing situation, the first and the last volumes in particular
containing long essayistic passages reflecting on life, death, art and
literature. In long sections he changes the focalization and even the
narrative tense from past to historical present, so that it isn’t the narrating
I’s but the experiencing I’s (Karl Ove’s) perspective we are confronted
with. In these chapters there are numerous examples of the kind of
unreliability Phelan focused on in McCourt’s case, with the difference,
however, that unreliability is established here due to the distance between
a diegetic narrator and a fallible filter.
But these examples are less interesting in the present perspective. It is
more interesting that there are aspects of the work which raise doubts
about and even undermine the author-narrator’s reliability.

3
The first volumes have been published in several languages including German,
English and French. Due to the obvious allusion to Adolf Hitler’s Mein Kampf (an
issue to which attention is given in the sixth volume of the series), the books have
appeared under other titles than in Scandinavia. The first volume is called Sterben
(2011) in German, La Mort d’un père (2012) in French and A Death in the Family
(2012) in English. In the latter case, the series has been called My Struggle. In the
following pages I will refer to the series name.
 Autofiction and Authorial Unreliable Narration 7

In volume two, in the midst of a very detailed retelling of a long

conversation during a dinner party at a restaurant, Karl Ove and his friend
Geir agree upon the fact that Karl Ove has always had an extremely bad
memory. Due to the frame, the statement becomes a variant of
Epimenides’ paradox: “All Cretans are liars.” An author giving a
hyperdetailed account of a long-past situation cannot in the same situation
claim to have a bad memory. The impact of this inconsistency is quite
remarkable. The conversation between Geir and Karl Ove is on a very
highbrow intellectual level of almost Socratic-dialogic dimensions,
focusing on, among other things, the personal costs connected with “the
authorial call,” etc. But the remark concerning Karl Ove’s weak memory
makes the reader speculate whether the whole dialogue isn’t more a
matter of the authorial I’s creative attempt to put himself in a better light,
due to the decisions regarding family life and relationships he has had to
make to pursue his career as an author. We begin speculating whether
Knausgård-the-author’s reporting is reliable.
Another example: over the course of the six volumes, Knausgård
returns to some of the significant incidents in his life. Among these is his
complicated relationship to his tyrannical and choleric father who died of
alcohol abuse ten years prior to the time of writing. The last period of the
father’s life, as it was experienced by Karl Ove, is focused on in both the
first and the last volumes. But in the latter case, the description is much
more downtoned than in the former. And where Karl Ove and his older
brother Yngve were pretty much alone in cleaning up the mess after the
father in the first volume, they receive attention and help from an uncle
and aunt in volume 5. Both descriptions focalize the incidents through the
experiencing I, but it seems as if the narrating I—the author—through the
process of writing the novel, develops his understanding of the story as it
is told, a contradiction that exposes a dynamic unreliability regarding
values and judgments. In volume 6, Knausgård receives an angry letter
from his uncle (his father’s brother) in which the author’s version of the
incidents in relation to the father’s death is being questioned with
reference to supposed hard evidence. And even though Knausgård
becomes aware that there might be flaws and misjudgments in his former
retelling of the incidents, he is not abandoning it. As he says to a
prosecutor in an imagined trial: “This is how I remember it” (volume 6,
303; translation mine). But as we (and he himself) know: he has an
extremely bad memory.
The death of the father is not the only example of this kind of
unreliability. In volume 1 we are told that Knausgård spent four years
8 Per Krogh Hansen

with a girl he didn’t love; but when the relationship is described for us in
the fifth volume, it most certainly seems characterized by love. Again: the
authorial perspective on and understanding of the incidents has taken
over, and the result is that our natural urge to rely on the authorial voice
of the narrative is challenged. Where our expectation with regard to the
authorial narrative agent is that it is stable and sanctions the norms and
values of the storyworld, we are instead engaging with inconsistencies,
and the result for our reading is that we redirect our attention from the
told toward the teller and thus see the misrepresentations and
inconsistencies as an expression of character traits and unreliable
narration. In that sense, Knausgård is, either intentionally or
unintentionally, flouting the assumption of a stable author. In the
beginning, we greet the representation as reliable. But as the telling and
retelling progress, we start having second thoughts, just as Knausgård
himself does in the sixth volume, when (as pointed out above) he refers to
memory instead of fact.
The examples commented on so far are all intra- or internarrational
insofar as the effect of unreliability is a result of contradictions within the
narrator’s discourse or between his discourse and the differing perspective
of others (e.g., his uncle’s) perspective on the same incidents.
But there are also examples of extratextual circumstances influencing
the author-narrator’s reliability.4 Even though the books were published
as “novels”, Knausgård claimed that everything he told was true. All
material had been presented to the persons concerned, and only a few
names were changed at their request.
Everything in this work is depicted with an attention to detail that
marks the work as a fictional recreation of the past: no one can remember
their past as clearly and elaborately as Knausgård does. And even though
we as readers might accept the level of detail, bearing in mind that this is
also a work of fiction, the authenticity is punctured from inside the
storyworld by the aforementioned unreliability signals.
Due to the work’s claim to factuality, moreover, it was also punctured
from the outside by the persons and family members depicted. Several felt
exposed and misrepresented and expressed their displeasure and
disappointment through the media. Fourteen family members even

4
The concepts intranarrational, internarrational and extratextual unreliability belong,
together with intertextual unreliability, to a taxonomy of different ways of
signaling/detecting unreliable narration. For further detail, see Hansen (2005, 2007,
2009).
 Autofiction and Authorial Unreliable Narration 9

announced that they wanted to bring Knausgård to trial for his “Judas
literature.”5
Now, Lejeune claimed that when an author writes in his own name, he
signs an autobiographical pact in which everything stated is to be
considered true. If something turns out to be false, it is not a matter of
fictionalization but, as quoted earlier, relates to the “order of lying.” This
rather strict understanding of autobiography’s truth value has been revised
in later conceptions of the genre. As Phelan notes, autobiographical
theory has repeatedly shown that “subjective truth is far more important
to memoir than literal truth […] because it is crucial to the
autobiographer’s ability to give shape and meaning to experience” (2005,
73). But as he also remarks, subjective “truth must also be accountable to
some extent to facts, people, and events that have an existence
independent of the autobiographer’s perception” (73).
With reference to Lejeune’s initial distinction, it can be observed that a
radical subjective recounting of the story jeopardizes the author’s
reliability and makes us, the readers, react in the same way as when we
get suspicious about a narrator’s account in fictional narrative: we read
with precaution and look beyond the authorial representation of the facts;
we try to figure out the true facts of the case and use our observations to
construct a critical picture of the authorial self about which we make
ethical judgments. We might even press charges against this self due to
the fact that autobiography belongs to the system of factual
communication that is restricted by responsibilities and protection against
defamation.
It is also in this respect that we find the main source of the difference
between the everyday use of the concept of unreliability and its use within
the context of fictional narrative: where the two forms of unreliability
both invoke the reader’s ethical judgment, only the latter opens up for an
esthetic judgment.
Autofiction blurs this distinction. Insofar as Knausgård’s work is also
claimed to be fiction, the system of factual cultural interaction and
communication is partly suspended, and instead the rules governing
fiction takes over: whatever is depicted is fictionalized, and thus a
simulated reality is told from a given perspective. Considered an author of
a work of fiction, Knausgård cannot be blamed for the misrepresentation
5
Uncle Gunnar (which isn’t his real name) was among the most critical voices to the
project, notably in a commentary published in the newspaper Fædrelandsvennen (cf.
Kristensen 2011). Knausgård’s estranged ex-wife, Tonje Aursland, also retorted in
October 2010 in a radio documentary broadcast on NRK.
10 Per Krogh Hansen

of persons and events or for his exaggerated remembrance of details.

Under the auspices of novelistic fiction, this kind of re-creational
depiction of the past is expected and allowed.
This, of course, also has significance for the reader’s approach to the
authorial self. Due to the many misrepresentations in Knausgård’s
narrative, the ethical judgments passed on both the actions of the author
and the characters and the critical and questionable aspects of the author’s
telling are accompanied by esthetic judgments. The narrative is not read
as an author’s telling about his past, but as a refiguration of that past.
By making himself the author, the narrator and the protagonist of his
work and at the same time claiming the storyworld to be in accordance
with reality, Knausgård’s result is more or less doomed to be an act of
unreliable narration. Reliability is dependent on a perspective that can tell
the truth. This is why, on a general level, it makes sense to speak of
fictional truth, insofar as narrative fiction is the creation of a storyworld
from a given perspective. But reality as such does not support one truth
alone. Truth in reality is perpetually being renegotiated. When Knausgård
claims that he tells the truth, but does it within the framework of fiction,
he suspends the negotiable nature of factual truth. By doing so, he is both
hit and saved by what we could call the “kernel paradox” of autofiction:
he claims that what he tells is true, thus opening up the possibility of
negotiation, while at the same time claiming that what he says is fiction,
authorizing the truth value of the told. It is in this tension between author
and narrator that Knausgård’s unreliability comes into being. More
generally, autofiction as a genre must at the very least be suspected of
extratextual unreliability.

4. Auto-Mocking and Self-Deception

From the perspective of unreliable narration as narration “in which the

narrator’s reporting, reading (or interpreting), and/or regarding (or
evaluating) are not in accord with the implied author’s” (Phelan 2005,
219), autofictional authorial unreliability of the kind we find in
Knausgård’s novels will clearly fall outside the definition. We could
therefore choose to invent a new term to cope with these deviant cases.6
But insofar as there are quite a few common aspects between the two

6
Phelan (2011) has suggested that we should distinguish between unreliable and
deficient narration, the latter covering an example such as the one I have discussed
here.
 Autofiction and Authorial Unreliable Narration 11

sorts of unreliability, I would like to avoid throwing yet another narrative

term on the heap of narratological concepts and models. Let me instead
close this article by considering whether room can be found within the
standard conception that will allow us to accommodate narrative works
such as Knausgård’s.
The concepts that serve as the toolbox we label “narratology” come
from a great variety of disciplines and discourses and are brought to
foreign areas thanks to the fact that someone has found an item that lacks
description but shares significant aspects with another item that has
already been described. The reason for this travelling applicability is the
fact that narrative is a transtextual, transgeneric and transmedial
phenomenon. Applying general narrative conceps developed in one
context to comparable items in another context gives us the added
advantage of providing our analysis with considering whether the
conception has been biased by the fact that it is formulated within a
framework of one genre, mediality or communicational mode.
Scholars from literary studies in particular have provided narratology
with a nearly endless series of useful concepts for the transmedial and
transdisciplinary study of narrative. But they have also quite often made
themselves guilty of making general claims as to what narrative is and
does, whereas what they are actually describing are specific features for
prose fiction, not narrative in general.
The concepts of unreliable narrator and unreliable narration have
travelled between genres and modalities ever since Wayne Booth coined
the term more than fifty years ago in his study of the rhetoric of novelistic
narrative fiction. Initially, at least, he did not consider whether the
concept was of general relevance or limited to literary fiction. Even so, it
seems that Booth did have a vague idea that there was more to the story
than he told.
Booth has often been blamed for the inconsistency of his initial
formulation:

For lack of better terms, I have called a narrator reliable when he speaks for
or acts in accordance with the norms of the work (which is to say, the implied
author’s norms), unreliable when he does not. (Booth 1991 [1961/83]: 158–
159)

All discussions of the concept since then have their basis in differences of
accentuation in this definition. Rhetoricians working on literary narrative
have focused on the parenthetically mentioned authorial agent,
12 Per Krogh Hansen

intentionality and normativity (understood as ethics) whereas more

semiotically oriented approaches have highlighted the lack of accordance
between the narrator’s representation of the storyworld and the actual
constitution of it. One might claim that the former has developed the
Boothian concept with respect to the fact that he was working with
literary, fictional narrative, whereas the latter have tried to include it in
more transdisciplinary approaches.
If we can agree that the concept has proven its value as a traveling
concept, I think we owe it ourselves to reconsider Booth’s definition and
liberate the “narrative/narratological” part of the concept from the literary
part.
By focusing on unreliabilty in Knausgård’s autofictional case, it is
clear that the implied author component is not a necessary part of the
unreliable narrator machinery. Rather, it is a concept that proves helpful
in the distinct cases of unreliable narration in which it is used as technique
for what Phelan calls “indirection.” This, we could claim, is a feature
related to fiction, a matter of fictionality, and thus important to study in
the study of narrative fiction.
With regard to the reliability of the narrator, seeking explanatory
recourse in the implied author is a matter of framing. What relates
unreliable narrators across the fact/fiction distinction, across genres and
modalities, is that they do not speak or act in accordance with the norms,
values or facts of the storyworld. As discussed above, it is characteristic
for fictional storyworlds that their constitution can be authorized by an
authorial agent. In factual narratives, by contrast, author and narrator are
often (as Cohn stated) the same. This, however, does not rule out the
possibility of unreliability. The storyworld is simply not governed by an
implied author in these cases, but rather by sensus communis to the extent
that readers have a stake in it.
Autofiction is a special case, since it blurs the borders between author,
narrator and character as well as between fact and fiction. In itself, it is an
extremely unreliable genre, and it can thus take either the form of
intentional auto-mocking, as in McCourt’s example, or that of seemingly
unintentional self-deception, as in Knausgård’s. So perhaps, after all,
there is still room and a need for new conceptions and distinctions.

Works Cited
 Autofiction and Authorial Unreliable Narration 13

Behrendt, Poul. 2006. Dobbeltkontrakten. En æstetisk nydannelse. København:

Gyldendal.
Booth, Wayne C. 1991 [1983/61]. The Rhetoric of Fiction. London, etc.: Penguin
Books.
Cohn, Dorrit. 2000. “Discordant Narration.” Style 34 (2): 307–316.
Doubrovsky, Serge. 1977. Fils. Paris: Galilée.
Genette, Gérard. 1993 [1991]. Fiction and Diction. Translated by Catherine
Porter. Ithaca, NY: Cornell University Press.
Hansen, Per Krogh. 2005. “When Facts Become Fiction: On Extra-Textual
Unreliable Narration.” In Fact and Fiction in Narrative: An Interdisciplinary
Approach, edited by Lars-Åke Skalin, 227–307. Örebro: Örebro University.
———. 2007. “Reconsidering the Unreliable Narrator.” Semiotica 165 (1/4):
227–246.
———. 2009. “Unreliable Narration in Cinema. Facing the Cognitive Challenge
Arising from Literary Studies.” Amsterdam International Elecronic Journal
for Cultural Narratology (ACJN) 5.
http://cf.hum.uva.nl/narratology/a09_hansen.htm
(Accessed 15 February 2016)
Knausgård, Karl Ove. 2009–2011. Min kamp 1-6. København: Lindhardt og
Ringhof.
Kristensen, Eivind. 2011. “‘Onkel Gunnar’ tar knallhardt oppgjør med
Knausgård.” Aftenposten. 17 February 2011.
http://www.aftenposten.no/kultur/Onkel-Gunnar-tar-knallhardt-oppgjor-med-
Knausgard-6699655.html (Accessed 15 February 2016)
Lejeune, Philippe. 1989 [1975]. On Autobiography. Translated by Katherine
Leary. Minneapolis: University of Minnesota Press.
McCourt, Frank. 1996. Angela’s Ashes: A Memoir. New York: Scribner.
Nünning, Ansgar. 1997. “‘But why will you say that I am mad?’ On the Theory,
History, and Signals of Unreliable Narration in British Fiction.” AAA.
Arbeiten aus Anglistik und Amerikanistik 22 (1): 83–105.
———. 1998a. “Unreliable Narration zur Einführung. Grundzüge einer
kognitiv-narratologischen Theorie und Analyse unglaubwürdigen Erzählens.”
In Unreliable Narration: Studien sur Theorie und Praxis unglaubwürdigen
Erzählens in der englischsprachigen Erzähllitteratur, edited by Ansgar
Nünning, Carola Surkamp and Bruno Zerweck, 3–40. Trier: WVT.
———. 1998b.”Unreliable Narrator.” In Encyclopedia of the Novel, edited by
Paul E. Schellinger, vol. 2: 1386–1388. Chicago and London: Fitzroy
Dearborn Publishers.
———. 1999. “Unreliable, compared to what? Towards a Cognitive Theory of
‘Unreliable Narration’: Prolegomena and Hypotheses.” In
Grenzüberschreitungen: Narratologien im Kontext/Transcending
Boundaries: Narratology in Context, edited by Walter Grünzweig and
Andreas Solbach, 53–73. Tübingen: Gunter Narr Verlag.
14 Per Krogh Hansen

Phelan, James. 2005. Living to Tell about It. A Rhetoric and Ethics of Character
Narration. Ithaca, NY: Cornell University Press.
———. 2011. “The Implied Author, Deficient Narration, and Nonfiction
Narrative: Or, What’s Off-Kilter in The Year of Magical Thinking and The
Diving Bell and the Butterfly?” Style 45 (1): 127–145.
Walsh, Richard. 2007. The Rhetoric of Fictionality: Narrative Theory and the
Idea of Fiction. Columbus: The Ohio State University Press.