As odd as it may seem in this first quarter of the 21st century, literary theory still hasn’t solved the real/fiction riddle.1 What does the real/fiction riddle consist of? It refers to the fact that even though many theories have been put forward, there is no consensus, and that’s an understatement, regarding what happens when two ontologies meet in a text defined by an apparently univocal genre and horizon of expectation. The use here of the verb “happen” is essential as it directly refers to the phenomenology of reading, of how readers make sense, consciously or not, of this ontological hybridity. One of the reasons this is still an ongoing issue stems from the fact that to resort to phenomenology implies calling on cognitive sciences and on this specific matter of texts comprising dual ontologies, resources in the field are logically very limited.
The fact that two of the most prominent concepts emerging from Narrative Studies in the last three decades are more or less directly related to the real/fiction relation is quite revealing of the sway it holds on this domain. Indeed, whether it is Monika Fludernik’s concept of Natural Narratology or Henrik Skov Nielsen, James Phelan and Richard Walsh’s theory of fictionality, there is in both cases a strong emphasis on how fiction and reality keep interacting and impacting each other. In her seminal Towards a Natural Narratology, the German scholar, drawing on Ricœur and White’s works which gave rise to the narrative turn, demonstrates how the very notion of narrativity is entangled in the way we think, and thus in the way we apprehend our physical and social environment (“situations tend to be comprehended holistically on the lines of frames and scripts” 17), while reciprocally, narratives are the very fruit of our experience, or of “experientiality”. She writes: “‘Natural’ narratology, as I envisage it, relies on a definition of narrativity as mediated human experientiality (which can be plotted on the level of action or on the level of fictional consciousness)” (36). She goes on to underline again how consciousness and narrativity heavily rely on each other: “Narrative is a category of human behaviour that occupies a very special place among behavioural modes. Its function seems to serve a deep-seated human need to cognize oneself in projected semiotic models of a narrative cast” (36). Quite distinctly at first sight, but with a similar underlying paradigmatic logic (i.e. how fiction and more generally narratives permeate through the real, and vice-versa), Nielsen, Phelan and Walsh present in “Ten Theses about Fictionality” fictionality as a key cognitive process in our daily lives. According to them,
[s]tories presented as invented are regularly and pervasively employed in political rhetoric; as vehicles of cultural memory and ideological negotiation of past and present; in thought experiments, scenario thinking, and risk assessments; and in many other areas of the societal, political, and cultural field. (63)
In other terms, fictional narratives are necessary tools as we carry on with our daily lives. These lives are not only about what actually happens but also about what might happen as we constantly elaborate counterfactuals in order to take the right decisions whatever the stakes are:
Human beings are concerned not only with matters of fact and with what is the case but also with evaluative questions that encompass possibilities and alternatives—with what is not the case and could never be the case, with what is not the case but could be the case, with what should have been the case, and so on. (64)
Natural narratology and fictionality epitomize a more general trend that can best be described by the will to take narrative studies beyond the domain of literature and show how narrativity is applicable to all aspects of human lives.
These two concepts also emphasize how major figures of these narrative studies have stopped, at least temporarily, devoting their attention to the study of how reality infiltrates fiction and have moved on instead to analyzing how the latter invades the former (although this is less true as far as Fludernik is concerned). This is all very interesting, but we are not any closer to a better understanding of how the two ontologies “cohabitate” within the same fictional space. And yet, this “cohabitation” has been a common feature of novels, and more particularly postmodern novels for quite a while now. This even pushed Linda Hutcheon to note that “[t]he most radical boundaries crossed […] have been those between fiction and non-fiction and—by extension—between art and life” (10). She typically and rightfully takes autofiction as an example since this must be the genre which has the most systematically tried to blur the lines between reality and fiction, perpetuating and above all intensifying a tradition initiated by the autobiographical novel: “challeng[ing] the life/art borders” and “play[ing] on the margins of the genre” (10).
Of course, some steps have been taken to deepen our understanding of the very nature of fiction, from a phenomenological (Jean-Marie Schaeffer’s Pourquoi la fiction?) or a stylistic and rhetorical perspective (Dorrit Cohn’s The Distinction of Fiction), or from both (Françoise Lavocat’s Fait et fiction). But these studies primarily focus on the essential distinction between fact and fiction, not on what happens when these two ontologies meet within the same text. The phenomenological aspect is key as it is important to bear in mind that reading is both a mental and physical activity. Richard Saint-Gelais summed up these two simultaneous levels of experience, these “two modes of apprehension, one material, the other fictional” thus: “on the one hand, turning the page and searching for the first line of the next chapter; on the other, the almost imperceptible suspension of and resumed attention to the characters and their imaginary actions and temporal and spatial settings” (57). Contrary to Cohn who really restricts her analysis to what fiction can do that referential writing cannot, Schaeffer through his concept of “immersion” takes the “two modes of apprehension” into account. However, Pourquoi la fiction ? as its title indicates is first and foremost about what characterizes fiction and our reading of it. Lavocat’s research embraces the full spectrum of what we commonly regard as fiction, a fluctuating notion, especially as “[t]he unstable status of fiction shows clearly that the definition of its boundaries has societal and political stakes”.2 She devotes a great part of her book to the panfictionalist approach (see 59–115), an extreme offshoot of the narrative turn according to which every narrative attempt is fictional, even if it aims at recounting facts and real events. Furthermore, “she points out that when it comes to fiction and the fact/fiction polarity, there is no single consensus within the scientific community, far from it” (Schmitt, Phenomenology 51) and eventually considers fiction “as a game of crossing ontological borders” (Schmitt, Phenomenology 61). However, even if she openly and quite extensively addresses the issue mentioned at the beginning of this article, she does this mostly from the perspective of fiction and her focus remains on redefining fiction as opposed to factual narratives. Her book, similarly to Cohn’s3 and Schaeffer’s, is an important contribution to fictionality studies, but while providing a very useful background to this study, it does not fully answer this article’s programmatic approach: what happens when Napoleon meets a fictional character in a novel? And by extension, what does Napoleon do in a novel?
Why Napoleon? Because in Possible Worlds, Artificial Intelligence and Narrative Theory, Marie-Laure Ryan famously used this example to illustrate a particular case in her extensive study of transuniverse relations: “a world in which Napoleon dies on St. Helena and successfully escapes to New Orleans is not possible, since it entails ‘Napoleon did and did not die on St. Helena.’ But there is nothing inconsistent about either one of these facts taken individually” (31). “There is nothing inconsistent” about the “fact” that Napoleon did not die on St. Helena from a fictional perspective of course. From a historical, thus a referential perspective, it is a different matter and many historians would argue that this “fact” is fully inconsistent with historical records. Additionally, the fact that Napoleon did or did not die on St. Helena radically impacts the perception readers have of this historical figure within a particular fictional context: A) he did die on St. Helena and readers will continue to regard the textual Napoleon as in keeping with the historical Napoleon, an extraneous referential figure imported into a fictional environment. B) he did not die on St. Helena and as the fiction proceeds on, textual Napoleon and historical Napoleon will drift apart, the former losing its referential features and more or less becoming a character like any other. Of course, doing whatever one wants to do with Napoleon is part and parcel of what we commonly call poetic license, although, as demonstrated by the numerous legal cases spawned by ambiguous memoirs or autofictional texts, or more generally what Françoise Lavocat called “imperfect fictionality or abusive factuality” (“fictionnalité imparfaite ou factualité abusive” 282), this very license often faces serious obstacles in its handling of living persons. There is a fundamental difference between Ryan’s case study and Emmanuel Carrère’s ex-wife who refused to be included in his autobiographical text Yoga,4 and between these two extremes lie many degrees of textualization of real human beings. John R. Searle takes a similar example to illustrate the difference between fiction and non‑fiction. Focusing on a short quote from Iris Murdoch’s the Red and the Green and comparing it to an extract from a New York Times article, Searle shows that despite rhetorical similarities, the author of the article is expected to comply with a certain number of rules, the most important criterion being whether her “assertion is defective” (12) or not. On the other hand, Murdoch’s utterance
is not a commitment to the truth of the proposition that on a sunny Sunday afternoon in April of nineteen-sixteen a recently commissioned lieutenant of an outfit called the King Edwards Horse named Andrew Chase-White pottered in his garden and thought that he was going to have ten more glorious days without horses. (12)
The novelist’s assertions are not supposed to bear any relation to the “truth”, they have no truth value, even in the framework of Realism.5 Of course, including real persons (dead or alive) is set to undermine, or at least question the novelist’s sense of creative freedom. It creates a problematic bridge between fiction and reality, one that entails responsibilities and a potential non-congruent response from some readers such as Emmanuel Carrère’s ex‑wife.
However, I am not interested in the legal dimension of this response and will “only” concentrate on how the reader fits this extraneous element within his experience of “playful, shared pretense” (“feintise partagée ludique”, Schaeffer 102). Furthermore, as underlined by Marie-Laure Ryan, contrary to what happens in phenomenal reality, there can be several versions of Napoleon in a novel, or more exactly any novelist can create her/his own version of the historical figure and most importantly, can choose to tamper with the historical facts and present readers with a much-altered version of a historical figure, which can be quite common in a uchronic novel. However, as altered as this figure can be, it is very often his/her “fate” that is changed whereas the psychological contours remain close to what is known about her/him, such as Charles Lindbergh in Philip Roth’s The Plot Against America. I will only consider the most faithful versions of these figures as they are the ones standing in stark contrast with their fictional environment and seemingly setting the two ontologies against each other.
Obviously, any realistic fiction, even more so any historical fiction (a term much more problematic than its common acceptance implies) tries to combine these two ontologies. In 1967, F. E. Sparshott, in his article entitled “Truth in Fiction”, took the example of Napoleon before Marie-Laure Ryan did:
Take first an extreme case, that of a historical novel. What is it that we are asked to imagine? Not that there was a man called Napoleon, for we know that there really was; nor that he escaped from Elba, for we know that he really did. Rather, we are asked to suppose that certain events in this well-known career took place in certain ways, although we do not know whether they in fact happened so or otherwise, or we may even know that they happened otherwise; or to suppose, that among these known episodes, others were interspersed that we know did not in fact take place. The world that we imagine contrives to be a world, and is the imaginary world it is, only because we supplement what the author tells us with what we remember of the actual world in which the real Napoleon lived. (4)
Several elements in this quote echo what has been developed above. First, more than fifty years ago, Sparshott asked a similar question to the one I am asking in this article, in a very different theoretical context: What is it that we are asked to imagine when Napoleon pops up in a novel? “What are we asked to imagine” is quite similar to “what type of experience does it create?” Put differently, he took into consideration the phenomenological dimension of this type of hybridity at a time when such considerations were highly uncommon. Then, he investigates the modus operandi of the author’s poetic license in historical novels and points out how this genre is a mix of fictional content and of information drawn from the readers’ semantic memory (“only because we supplement what the author tells us with what we remember of the actual world in which the real Napoleon lived”). Although his article doesn’t offer any concrete solution regarding how fictional and referential input cohabit and interact within this mostly fictional world, he had the merit of raising the issue and few theorists have picked up the gauntlet since then, quite surprisingly since the “issue” remains highly topical as historical novels have turned into Biofiction, a booming genre.
I would like to expand a little bit more on Sparshott’s article, and focus on another quote, again relevant to the real/fiction dichotomy:
The type of fiction most current is not the historical novel, but the novel in which a fictional character is placed in a familiar setting. We are not then asked to suppose (or imagine) that there is a place called London, for we know there is; nor that there is a Baker Street in London, for we may know that too. And we may well know without being told whereabouts in London Baker Street is; in fact, our full comprehension of the fiction depends on this knowledge. What we are asked to suppose is that in Baker Street there is a No. 221b (of which most of us do not know whether there is or not, although we do know where it would be if it existed), and that at No. 221b a Mr. Sherlock Holmes has lodgings (which we are quite sure is not the case). We are asked, in fact, to imagine that among the people we know move others we do not know, that among the streets with which we are familiar are others with which we are not familiar: we are invited to imagine in familiar places and their populations those changes, and only those changes, that the author postulates. If we make further changes, we have not succeeded in following the story that the author is telling. (4)
While, as mentioned before, Sparshott also turns his attention to historical novels and their potentially hybrid dimension, the “piece of reality” imported in the novel he surprisingly takes as an example here is not a historical figure or event but the type of topographical information almost any more or less realistic novel includes, historical or not, London and Baker Street in this specific case. First, there is a fundamental difference between the representation of an actual city you have visited or even lived in and of one you have never set foot in. In the first case, you rely on your episodic memory, things you have actually seen or experienced, in the second case, you use your semantic memory, things you have learned. But there sometimes is a thin line between the two as a city you have seen represented in a book or even more so in a film can become part of your experience, hence the familiar feeling you have when you first go to New York, a city that is the setting of a substantial number of visual fictional works. Furthermore, Sparshott does not take into account the fact that mental images based on literary mimetic descriptions are highly unstable phenomena, and yet this has a tremendous impact on the way we visualize textual topographical information. In What We See When We Read, Peter Mendelsund writes that “[w]hat we do not see is what the author pictured when writing a particular book” (207) and claims that “[w]e take in as much of the author’s world as we can, and mix this material with our own in the alembic of our reading minds, combining them to alchemize something unique” (207). In fact, Mendelsund undermines one of the most fundamental aspects of realism, its ability to recreate, or at least evoke the actual world by supposedly triggering off the corresponding visual imagery. Taking the example of Dickens’s description of an industrial harbor in Our Mutual Friend, he demonstrates that mental representations of texts are ruled by a form of visual determinism and that in fact what we see when we read is what we have already seen when we don’t read. Finally, it is important to note that even in the case of auto/biographical writings, what Ryan calls the “‘language game’ of nonfiction is not defined by the objective relation of the text to the world, but by the rules that govern the use of the text, and bind sender and receiver in a communicative contract” (Ryan, “Postmodern” 166). But the representation of the real is too wide a topic for this article and again I only want to deal with one specific aspect of it. Sparshott’s long quote is nevertheless interesting first because it reveals that the mimetic process in historical fiction is far from simple, and above all that the inclusion of historical figures in a fictional context is the exact opposite of what he describes when we read a Sherlock Holmes story: “We are asked, in fact, to imagine that among the people we know move others we do not know, that among the streets with which we are familiar are others with which we are not familiar.” As a matter of fact, if we read a story in which Napoleon is described as taking part in events that never took place along characters who never existed, we are asked to imagine that among the people we do not know, move others that we know or at least know really existed.
But as opposed to Baker Street, Napoleon is not supposed to generate mental images but to conjure up historical as well as psychological content. This poses the problem of the nature of this fictional Napoleon and the type of space “he” creates within a fictional universe. The term “fictional Napoleon” can be interpreted in two very different ways: it can either be seen as a version of Napoleon that remains close to the known biographical data, the Napoleon almost everyone has a certain amount of knowledge about, or as a version that is literally fictional, not simply because it is included in a novel such as the first version but because it is hardly recognizable on account of his being transformed by the novelist’s freewheeling imagination. As it turns out, there is always a bit of both as soon as you start imagining the historical figure’s psychological life even if the latter fits what we know about this person’s life. For instance, we know that Lincoln was greatly affected by the loss of his 11‑year‑old son Willie, but what George Saunders describes in Lincoln in the Bardo is a hyperbolic version of known facts. Very similarly, to take an example from another domain, the Lenny Bruce featuring in the TV series The Marvelous Mrs. Maisel is cast in a historical mold so to speak, but because he becomes one of the main characters, or at least a recurring one, he is bound to depart from this mold and become something more than what we know about the nonfictional Bruce, something different. To put it differently, you never really know what goes on in a person’s mind, as documented as her/his life might be and once you start delving into this psyche, you inevitably drift away from the historical and move deeper into the fictional.
At the very core of fiction, historical or not, lies creative freedom and its founding principle is that it “makes no claim to external truth, but rather, guarantees its own truth” and “[s]ince the text of fiction creates its own world, it is the sole mode of access to it. Unlike texts of nonfiction, fictional texts do not share their reference world with other texts” (Ryan, “Postmodernism” 167). But Ryan reminds us of another, even more important rule:
Though the fictional text creates an autonomous fictional world, this world can present some degree of overlap with the real world. […] In historical novels, many of the propositions expressed by the fictional discourse happen to be true in reality; but this does not turn the text into a blend of fiction and non fiction. The primary reference world is the fictional world, and unless the narrator is unreliable, all the propositions expressed by the text yield truths about this world. But because propositions can be separately valued in different possible worlds, the real world may function as secondary referent. (“Postmodern” 167–68)
This is indeed an important rule, but a problematic one. Napoleon is an obvious case of “overlap”, or at least of the “secondary referent” foraying into the “primary reference world”, but as interesting as Ryan’s observations may be, they do not tell us what the real outcome of this overlapping phenomenon is. Does the secondary reference world, as isolated and limited as its presence can be in a novel, retain its ontological integrity? In other words, does “Napoleon” remain Napoleon? Moreover, Ryan omits to differentiate between the multifarious ways the real world’s presence can be felt in a novel: a reference to a person is not the same as a reference to a place, which is not the same as a reference to an event and so on. And of course, there is the obvious example of the metalepsis of the author, when authors pop up in their own fiction, The French Lieutenant’s Woman being an obvious example. In the final analysis, this trope is nothing more than, again, the secondary reference world suddenly coming up to the fictional surface. Is the “John Fowles” in this novel the John Fowles? There is no better answer to this question than Brian McHale’s:
Intended to establish an absolute level of reality, it paradoxically relativizes reality; intended to provide an ontologically stable foothold, it only destabilizes ontology further. For the metafictional gesture of sacrificing an illusory reality to a higher, ‘realer’ reality, that of the author, sets a precedent: why should this gesture not be repeatable? What prevents the author’s reality from being treated in its turn as an illusion to be shattered? Nothing whatsoever, and so the supposedly absolute reality of the author becomes just another level of fiction, and the real world retreats to a further remove. (197)
Strikingly enough, Dorrit Cohn offers a similar analysis:
To me these ambiguous texts [ambiguous autobiographical novels for instance] indicate […] that we cannot conceive of any given text as more or less fictional, more or less factual, but that we read it one key or the other—that fiction, in short, is not a matter of degree but of kind, in first- no less than in third-person form. I hold to this position even in the face of the work that appears to challenge it most powerfully, a work Harry Levin called “the most extensive exercise in the first-person singular”. I mean, of course, À la recherche du temps perdu. (35–36).
Thus, according to McHale or Cohn, there is no “overlapping”, no foraying, no co-presence but simply a very no-nonsense either/or situation. Cohn directly contradicts Ryan according to whom “[i]f objects are inherently fictional or real, how can one explain the presence of historical individuals and real locations in a work of fiction?” (Possible Worlds 15). But Ryan fails to provide us with a clear notion of what the nature of this “presence” really is, which undermines her own analysis. Cohn also affirms that “fiction, in short, is not a matter of degree but of kind” which also contradicts Kai Mikkonen’s view that instead of seeing fiction and nonfiction as polar opposites, we should consider “the relation between fiction and factual representation” as “one of degree, a matter of a continuum of hybrid forms and thereby affected by the possibility of constant fluctuation between the two” (294–95). Of course, what these “degrees” or this “constant fluctuation” might be from the perspective of the reader is never developed, or even broached as a matter of fact.
Cognitive psychologist Richard J. Gerrig famously updated one of the oldest adages of literary theory, Coleridge’s “willing suspension of disbelief”, by claiming that when we read fiction, we are not skeptical, quite the opposite we tend to believe that everything is true or plausible. He reverses Coleridge’s truism and reveals that the real effort on the part of the reader is the “construction of disbelief” (240) which means that we do not need to willingly suspend disbelief since we naturally believe. Gerrig’s conclusions have several consequences when we read a generically ambiguous text:
-
Reading any book as fiction is similar to a default position.
-
It proves, if need be, that the nature of the text is determined by the position of the reader, and, much more often than not, “ambiguous” means fictional.
-
If we decide to disbelieve, it takes an additional cognitive effort. (Schmitt, Phenomenology 61)
How can Gerrig’s “construction of disbelief” contribute to solving the riddle of our fictional Napoleon? By keeping this concept in mind and calling on McHale’s interpretation of the mechanisms of metalepsis again, we may now wonder what happens when the reader does not suspend disbelief while reading a supposedly fictional text in which historical figures appear. McHale’s view is that once authors have entered their own fictional world, they are contaminated by the ontological environment and thus become fictional: “Fiction cancels the reality of the real, of the author. Reality is dissolved within fiction, which is in line with Gerrig’s construction of disbelief. This leads us to what I would call ‘The Purple Rose of Cairo illusion’” (Schmitt, Phenomenology 61). Why The Purple Rose of Cairo? The film’s plot hinges on transuniverse relations, as during the screening of a film, the main character, a cinema enthusiast, sees her favorite actor emerge from the black and white film she is watching, and literally from the screen, into her colorful, very real (to her, not to us viewers) world. Both worlds are logically impacted by this bold (and of course highly unlikely) move. Woody Allen’s narrative strategy is to build his plot on the very principle of immersion and entertainingly and metaphorically show how fiction can deeply affect our lives (“It is only human to mistake the make-believe for the factual or to believe a lie”, Mikkonen 293). Mikkonen playfully reminds us that fictional characters sometimes feel very real to us: “People also appropriate sayings, mannerisms, and behavior patterns from fiction, factual texts borrow narrative devices and metaphors from fiction, and sometimes real events are best described by figures of speech that refer to fiction” (293). This is undeniably true but this does not contradict McHale’s logic: if an element crosses an ontological border, the screen in this case, it irremediably is contaminated and the black and white character finds colors in his new reality. Of course, Allen’s character goes in the opposite direction of our fictional Napoleon: whereas the former moves from fiction to “reality”, the latter does exactly the opposite. And yet, it is my contention than in both cases, there can be no cohabitation, only contamination to use Cohn’s own very apt terminology, and one becomes the world one enters. For instance, to take again the example of how characters influence our own personalities or decisions, it is important to remember that these very influences, imported from fiction, are no longer fictional once they become either actionable behavior or part of our psychological profile. A more extreme case: if you start acting like a superhero and jump off a roof, there will soon be a painful reality check. It is in no way different in the other configuration, since when “historical persons […] interact with fictional characters”, they quite logically “perform actions that they could not have performed in ‘real life’, since these actions link them with invented characters. Hence, they are fictionalized”.6 Schaeffer’s “fictionalization” is quite similar to what I mean by “contamination”: the nature of the element imported is drastically altered, so much so that it bears little resemblance to the original version. Even the most genuine version of Napoleon—a highly questionable assertion as how can a man dead for several centuries be similar to his evocation in a novel published so longer after his demise?—will have to comply with the rules of the primary, and only reference world. Fictional Napoleon is not real Napoleon, at best he is a fictional ghost of the referential one.
We should also bear in mind that any hybrid proposition, whether tenable or not, requires a lot of attentional energy and that “once the reader has established a prevalent perspective, he tends to persevere with it as long as possible” (Fludernik 20). Indeed, “we normally avoid mental overload by dividing our tasks into multiple easy steps” and abide by “the law of least effort” (Kahneman 38). The principle of contamination is a direct consequence of this cognitive reality; to some extent we cognitively surrender to the dominant ontology and it is through the highly distorting prism of the latter that we perceive these exogenous elements, such as Napoleon, or more exactly “Napoleon”.
To conclude, there are by and large two schools of thought when it comes to the overlapping of fiction and nonfiction and they are very well summed up by Mikkonen:
Radical forms of segregationism think that fiction is pure imagination without truth, that fiction has no ontological status, while radical integrationism assumes that there is no genuine ontological difference between fiction and true representation […]. The segregationist ontology argues that fiction can put forward true statements—but does not need to do so—only about imaginary beings (the truth value of a proposition may only be assigned separately for each possible world). Fiction does not therefore equal lying but involves nonexisting or nonactualizable entities. (303)
I have adopted in this article a clear stance in favor of “segregationism” even if the type of segregationism I feel close to is not based on the belief that “fiction is pure imagination without truth, that fiction has no ontological status”. Fiction does have an “ontological status”, but one that is fundamentally distinct from reality. This difference conjures up Deleuze and Guattari’s brilliant observation that “art does not think less than philosophy, but it thinks through affects and percepts” (“L’art ne pense pas moins que la philosophie, mais il pense par affects et percepts” 64). I simply believe that each of these two ontologies “swallows up” and “contaminates” any element imported from the other ontology. It was my purpose to argue that fiction can actually involve existing or “actualizable” entities, such as Napoleon, but once the latter is integrated into its new environment, it loses its reality and becomes part and parcel of the immersion experience and of a new language game. To answer a question previously asked—what does Napoleon do in a novel?—, “Napoleon” does exactly what fictional characters do: it is one of the tools aimed at creating a fictional experience.
