Six Walks in the Fictional Woods

by Umberto Eco

  Readers would be justified if they started to suspect that the empirical author was Poe, who invented a fictitious real person, Mr. X, who speaks of a false real person, Mr. Pym, who in his turn acts as the narrator of a fictional story. The only embarrassing thing is that these fictitious persons speak of the real Mr. Poe as if he were an inhabitant of their fictitious universe (see Figure 4).

  Who is the model author in all this textual tangle? Whoever it is, it is the voice, or the strategy, which confounds the various presumed empirical authors, so that the model reader can’t help becoming enmeshed in such a catoptric trick.

  Let’s go back to rereading Sylvie. By using an imperfect tense at the beginning, the voice which we have decided to call Nerval tells us that we are to prepare ourselves for listening to a reminiscence. After four pages the voice immediately shifts from the imperfect to the simple past and recounts a night spent in the club after the theater. We are to understand that here, too, we are listening to one of the narrator’s reminiscences, but that he is now recalling a precise moment. It is the moment in which, while talking to a friend about the actress he has loved for some time, without ever approaching her, he realizes that what he loves is not a woman but an image (“I pursue an image, nothing more”). Now, however, in the reality fixed precisely by the past tense, he reads in a newspaper that, in reality, that very evening in Loisy, where he spent his childhood, they are holding the traditional festival of archers, which he used to take part in as a boy, when he was infatuated with the beautiful Sylvie.

  In the second chapter, the story returns to the imperfect. The narrator spends a few hours in a half-sleep in which he recalls a similar festival, presumably when he was a boy. He remembers gentle Sylvie, who loved him, and the beautiful, haughty Adrienne, who had sung that evening on the lawn; she had been an almost miraculous apparition, and had then disappeared forever behind the walls of a convent. Between sleeping and waking, the narrator wonders whether he still hopelessly loves the same image—that is, whether in some inexplicable way both Adrienne and the actress are the same woman.

  Figure 3

  Figure 4

  In the third chapter the narrator is seized by the desire to return to the scene of his childhood memories, calculates that he could arrive there before dawn, goes out, takes a carriage, and, during the journey, as he is beginning to make out the roads, hills, and villages of his childhood, begins a new reminiscence—this one more recent, only about three years before the time of his journey. But the reader is introduced to this new stream of memories by a sentence which, if we read it carefully, seems amazing:

  Pendant que la voiture monte les côtes, recomposons les souvenirs du temps où j’y venais si souvent.

  While the carriage climbs the slope, let us recollect the memories of the time I came there so often.

  Who is it who pronounces (or writes) this sentence and calls for our involvement? The narrator? But the narrator, who is describing a carriage ride made years before the moment in which he is narrating, should say something like, “While the carriage was going up the hills, I recollected [or “I started recollecting,” or “I told myself ‘let us recollect’”] the memories of the days when I used to go there so often.” Who is—or rather, who are—those “we” who together have to bring back the memories and therefore make preparations for another journey into the past? Who are the “we” who have to do it now, “while the carriage is climbing” (while the carriage is moving at the same time as we are reading), and not then, “when the carriage was going,” at the moment the narrator tells us that he was recollecting? This is not the voice of the narrator; it is the voice of Nerval, the model author, who for a moment speaks in the first person in the story and says to us model readers: “While the narrator is going up the hills in his carriage, let us recompose (with him, of course, but you and I, too) the memories of the time when he would come so often to these places.” This is not a monologue but one part of a dialogue among three parties: Nerval, who surreptitiously enters the narrator’s discourse; we, who have been called upon to participate just as surreptitiously, when we thought we could observe the event from the outside (we, who thought we had never left a theater); and the narrator, who is not excluded, since it was he who came to those places so often (“J’y venais si souvent,” “I used to go there so often”).

  Figure 5

  It should also be pointed out that many pages could be written about that “j’y.” Does it mean “there,” there where the narrator was that evening? Or does it mean “here,” here where Nerval is suddenly taking us?

  At this point, in this tale where time and place are so inextricably tangled, even the voices seem to be confounded. But this confusion is so admirably orchestrated that it is imperceptible—or almost, since we are perceiving it. It is not confusion but a moment of clearsightedness, an epiphany of storytelling, in which the three components of the narrative trinity—the model author, the narrator, and the reader—all appear together (see Figure 5). They must appear together because the model author and the model reader are entities that become clear to each other only in the process of reading, so that each one creates the other. I think this is true not only for narrative texts but for any sort of text.

  In his Philosophical Investigations (number 66), Wittgenstein writes:

  Consider for example the proceedings that we call “games”: I mean board-games, card-games, ball-games, Olympic games, and so on. What is common to them all?—Don’t say: “There must be something common, or they would not be called ‘games’”—but look and see whether there is anything common to all.—For if you look at them you will not see something that is common to all, but similarities, relationships, and a whole series of them at that.22

  The personal pronouns in this passage do not indicate an empirical character or an empirical reader; they are merely textual strategies, set out as a form of appeal, as in the beginning of a dialogue. The intervention of a speaking subject occurs simultaneously with the creation of a model reader who knows how to continue the game of inquiry into the nature of games; and the intellectual disposition of this reader (even the urge to play on the subject of playing games) is determined only by the type of interpretive moves that that voice asks him or her to make: to look, to see, to consider, to find relationships and similarities. In the same way, the author is but a textual strategy that is capable of establishing semantic correlations and that asks to be imitated: when this voice says “I mean,” it is inviting us to come to an agreement, so that the word “game” should be taken as referring to board games, card games, and so on. But this voice does not define the word “game”; rather, it asks us to define it, or to recognize that it can be defined satisfactorily only in terms of “family resemblances.” In this text, Wittgenstein is merely a philosophical style, and his model reader merely the will and ability to adapt to this style, cooperating to make it possible.

  So I, a voice without body or sex or any history—unless it be the voice which starts with this first lecture and will end with the last—invite you, gentle readers, to play my game with me during our next five meetings.

  TWO

  THE WOODS OF LOISY

  There are two ways of walking through a wood. The first is to try one or several routes (so as to get out of the wood as fast as possible, say, or to reach the house of grandmother, Tom Thumb, or Hansel and Gretel); the second is to walk so as to discover what the wood is like and find out why some paths are accessible and others are not. Similarly, there are two ways of going through a narrative text. Any such text is addressed, above all, to a model reader of the first level, who wants to know, quite rightly, how the story ends (whether Ahab will manage to capture the whale, or whether Leopold Bloom will meet Stephen Dedalus after coming across him a few times on the sixteenth of June 1904). But every text is also addressed to a model reader of the second level, who wonders what sort of reader that story would like him or her to become and who wants to discover precisely how the model aut
hor goes about serving as a guide for the reader. In order to know how a story ends, it is usually enough to read it once. In contrast, to identify the model author the text has to be read many times, and certain stories endlessly. Only when empirical readers have discovered the model author, and have understood (or merely begun to understand) what it wanted from them, will they become full-fledged model readers.

  Perhaps the text in which the voice of the model author appeals most explicitly for the second-level reader’s collaboration is a famous detective story: The Murder of Roger Ackroyd, by Agatha Christie. Everyone knows the story. A narrator, speaking in the first person, tells how Hercule Poirot gradually comes to discover the culprit, except that at the end we learn from Poirot that the culprit is the narrator, who cannot deny his guilt. But while he is waiting to be arrested and about to commit suicide, the narrator turns directly to his readers. Indeed, this narrator is an ambiguous figure because not only is he the character who says “I” in a book written by somebody else; he also appears as the man who has physically written what we are reading (like Arthur Gordon Pym), and at the end of the story he acts as the model author of his own diary—or, if you prefer, the model author speaks through him, or, better, we enjoy through him the narrative representation of a model author.

  The narrator, therefore, invites his readers to read the book again from the beginning because, he states, if they had been perceptive, they would have realized that he had never lied. At most he had been reticent, because a text is a lazy machine that expects a lot of collaboration from the reader. And he not only invites the reader to read it again but physically helps the second-level reader to do so by quoting at the end some of the sentences from the opening chapters.

  I am rather pleased with myself as a writer. What could be neater, for instance, than the following: “The letters were brought in at twenty minutes to nine. It was just on ten minutes to nine when I left him, the letter still unread. I hesitated with my hand on the door handle, looking back and wondering if there was anything I had left undone.” All true, you see. But suppose I had put a row of stars after the first sentence? Would somebody then have wondered what exactly happened in that blank ten minutes?

  At this point the narrator explains what he really did in those ten minutes. Then he continues:

  I must admit that it gave me rather a shock to run into Parker just outside the door. I have faithfully recorded that fact. Then later, when the body was discovered, and I had sent Parker to telephone for the police, what a judicious use of words: “I did what little had to be done”! It was quite a little—just to shove the dictaphone into my bag and push back the chair against the wall, in its proper place.1

  Of course model authors are not always so explicit. If we go back to Sylvie, for example, we find ourselves dealing with an author who perhaps didn’t want us to read the text again, or rather who wanted us to reread but did not want us to understand what had happened to us while reading it the first time. Indeed Proust, in the pages he dedicated to Nerval, describes the impressions that any of us would be likely to have after reading Sylvie for the first time:

  What we have here is one of those rainbow-painted pictures, never to be seen in real life, or even called up by words, but sometimes brought before us in a dream or called up by music. Sometimes, in the moment of falling asleep, we see them, and try to seize and define them. Then we wake up and they are gone . . . It is something vague and haunting, like a memory. It is atmospheric, the atmosphere of Sylvie, a colouring in the air like the bloom on a grape . . . But it is not in the words, it is not said, it is all among the words, like the morning mist at Chantilly.2

  The word “mist” is very important. Sylvie really does seem to affect its readers like a mist, as if we were looking at a landscape through half-closed eyes, without clearly distinguishing the shape of things. But it is not that the things cannot be distinguished; on the contrary, the descriptions of landscapes and people in Sylvie are very clear and precise, even of a neoclassical clarity. What in fact readers cannot grasp is where they are in time. As Georges Poulet said, Nerval’s past “plays ring-around-the-rosy with him.”3

  The fundamental mechanism in Sylvie is based on a continual alternation between flashbacks and flashforwards, and on certain groups of embedded flashbacks.

  When we are told a story which refers to a narrative time 1 (the time at which the narrated events occur, which may be two hours ago or a thousand years ago), both the narrator (in the first or third person) and the characters can refer to something that happened prior to the time of the narrated events. Or they can hint at something that, at the time of these events, has yet to occur and that is anticipated. As Gérard Genette says, a flashback seems to make up for something the author has forgotten, whereas the flashforward is a manifestation of narrative impatience.4

  Everyone uses such techniques when describing past events: “Hey, listen to this! Yesterday I met John—perhaps you remember, he’s the one who used to go jogging every morning two years ago [flashback]. Well, he was very pale, and I must admit I only realized later why [flashforward], and he says—oh! I forgot to tell you that when I saw him he was coming out of a bar, and it was only ten o’clock in the morning, get it? [flashback]—anyway, John tells me that—Oh, God, you’ll never guess what he told me [flashforward] . . .” I hope I won’t be so confusing in the rest of this discussion. But, with greater artistic sense, Nerval certainly confounds us throughout Sylvie with a dizzying game of flashbacks and flashforwards.

  The narrator of the story is in love with an actress, without knowing whether his love is reciprocated. An item in the newspaper suddenly brings back memories of his childhood. He returns home, and in a half-sleeping, half-waking state he remembers two girls, Sylvie and Adrienne. Adrienne was like a vision: blonde, beautiful, tall, and slender. She was “a mirage of glory and beauty”; “the blood of the Valois flowed in her veins.” Sylvie, in contrast, looked like a “petite fille” from the neighboring village, a country girl with dark eyes and “slightly tanned skin,” childishly jealous of the attention the narrator paid to Adrienne.

  After a few sleepless hours, the narrator decides to take a carriage and return to the place of his memories. During the journey he begins to recall other events (“recomposons les souvenirs . . .”)—events that occurred in a past closer to the time of the journey: “quelques années s’étaient ecoulées” (“some years had passed”). In this long flashback, Adrienne appears only fleetingly and as a memory within a memory, while Sylvie is strikingly alive and real. She is no longer a “little village girl . . . She had become so beautiful!” Her figure is lithe; she has something Athenian in her smile. She is now endowed with all the graces that the narrator in his youth attributed to Adrienne, and perhaps the narrator can satisfy his need for love with her. They pay a visit to Sylvie’s aunt, and, in a moving scene that seems to foretell their possible happiness, they dress up as fiancés from an earlier era. But it is too late, or too soon. The narrator returns to Paris the following day.

  Here he is now, as the carriage climbs a hill, bringing him back to his native village. It is four o’clock in the morning, and the narrator starts a new flashback, which we shall return to in another lecture—and please appreciate my flashforward, because here (in Chapter 7) the times really get completely mixed up and it is impossible to determine whether the last fleeting glimpse he had of Adrienne, which he only now remembers, was before or after the party he has just recalled. But the parenthesis is brief. We meet the narrator again on his arrival at Loisy, when an archery contest is drawing to a close, and he finds Sylvie once more. She is now a fascinating young woman, and the narrator recalls with her various aspects of their childhood and adolescence (the flashbacks occur in the story almost unnoticed); but he realizes that she, too, has changed. She has become a crafts-woman, producing gloves; she reads Rousseau, can sing arias from opera, has even learned “phrasing.” And finally, she is about to marry the narrator’s foster-brother and b
oyhood friend. The narrator realizes that the age of illusion cannot be recaptured and that he has lost his last chance.

  Back in Paris, the narrator finally succeeds in having a love affair with Aurélie, the actress. At this point the story speeds up: the narrator lives with the actress, understands that he doesn’t really love her, and returns with her a few times to the village, where Sylvie is now a happy mother; she is a friend and perhaps a sister to him. In the final chapter, the narrator, after being abandoned by the actress (or letting himself be abandoned), speaks once more to Sylvie, reflecting on his lost illusions.

  The story could be very banal, but the tangle of flashbacks and flashforwards makes it magically unreal. As Proust said, “One is constantly obliged to turn back to an earlier page to see where one is, if it is the present or the past recalled.”5 The misty effect is so pervasive that the reader usually fails in this task. It is clear why Proust, who was fascinated with the search for things past and who would end his work under the banner of time recaptured, considered Nerval both a master and an unsuccessful forerunner who lost his battle with time.

  But who is it who loses the battle? Gérard Labrunie, empirical author, destined to commit suicide? Nerval, model author? Or the reader? While he was writing Sylvie, Labrunie stayed several times in a clinic, in a rather critical state of mental health, and in Aurélia he tells us that he wrote laboriously, “almost always in pencil, on scattered bits of paper, following the haphazard course of my daydreams or my walks.” He wrote as an empirical reader at first reads, without seeing temporal ties, the before and after. Proust will say that Sylvie “is a dream of a dream,” but Labrunie really did write as if he were dreaming. This is not true of Nerval as model author. The apparent uncertainty concerning times and places which constitutes the fascination of Sylvie (and brings about a crisis in the first-level reader) is founded on a narrative strategy and grammatical tactics as perfect as clockwork—which, however, are visible only to the second-level reader.