Levels of Reality in Literature

The uses of literature - Italo Calvino 1986

Levels of Reality in Literature

Paper read at an international conference on “Levels of Reality,” Florence, September 1978.

Different levels of reality also exist in literature; in fact literature rests precisely on the distinction among various levels, and would be unthinkable without an awareness of this distinction. A work of literature might be defined as an operation carried out in the written language and involving several levels of reality at the same time. From this point of view, some consideration of works of literature might not be completely useless even to the scientist or philosopher of science.

In a work of literature, various levels of reality may meet while remaining distinct and separate, or else they may melt and mingle and knit together, achieving a harmony among their contradictions or else forming an explosive mixture. Shakespeare’s plays provide us with a number of clear examples. For distinction between the different levels we might think of A Midsummer Night’s Dream, in which the complications in the plot occur where three levels of reality intersect, though these remain quite distinct: (1) the aristocratic characters at the court of Theseus and Hippolyta; (2) the supernatural characters Titania, Oberon, and Puck; (3) the rustic comic characters, Bottom and his friends. This third level borders on the animal kingdom, which may be seen as a fourth level, entered by Bottom when he is changed into an ass. In addition there is one further level to consider, that of the performance of the story of Pyramus and Thisbe, the play within the play.

Hamlet, on the other hand, constitutes a sort of short circuit, or a whirlpool that sucks in all the various levels of reality; it is from their very irreconcilability that the drama comes into being. There is the ghost of Hamlet’s father with his demand for justice, which is the level of archaic values, of knightly virtues, with its moral code and supernatural beliefs; there is the level we might call realistic, that of “Something is rotten in the state of Denmark” (i.e., the court at Elsinore); there is the level of Hamlet’s inner life, of the modern psychological and intellectual awareness that is the great novelty of the play. To hold these three levels together, Hamlet disguises himself in a fourth, the linguistic barrier of his feigned madness. But, as if by induction, this feigned madness leads to real madness, and the level of madness seizes and eliminates one of the few positive elements remaining in the play: the delicate figure of Ophelia. This drama also has a play within the play, the performance of the troupe of strolling players; and this constitutes a level of reality on its own, separate from the rest, though interacting with them.

Up till now I have confined myself to distinguishing the various levels of reality within the work of art considered as a world of its own, but we cannot stop there. We have to consider the work as a product, in its relation to the outside world in the age when it was created and the age when we received it. In all periods and in all literatures we find works that at a certain time turn around on themselves, look at themselves in the act of coming into being, and become aware of the materials they are made of. Just to stick to Shakespeare, in the last act of Antony and Cleopatra, before killing herself, Cleopatra imagines her fate as a prisoner taken to Rome for Caesar’s triumph, mocked by the crowds; even now she thinks that her love for Antony will become the subject of theatrical performances:

… the quick comedians

Extemporally will stage us, and present

Our Alexandrian revels. Antony

Shall be brought drunken forth, and I shall see

Some squeaking Cleopatra boy my greatness

I’ the posture of a whore.

There is a fine passage by Middleton Murry about this dazzling piece of mental acrobatics. On the stage of the Globe theatre a piping boy dressed up as Cleopatra represents the real, majestic Queen Cleopatra in the act of imagining herself being represented by a boy dressed up as Cleopatra.

These are the tangles we have to start with in saying anything about the levels of reality in a work of literature. We cannot lose sight of the fact that these levels are part of the written world.

“I write.” This statement is the one and only real “datum” a writer can start from. “At this moment I am writing.” Which is also the same as saying: “You who are reading are obliged to believe only one thing: that what you are reading is something that at some previous time someone has written; what you are reading takes place in one particular world, that of the written word. It may be that likenesses can be established between the world of the written word and other worlds of experience, and that you will be called on to judge upon these likenesses, but your judgment would in any case be wrong if while reading you hoped to enter into a direct relationship with the experience of worlds other than that of the written word.” I have spoken here of “worlds of experience,” not of “levels of reality,” because within the world of the written word one can discern many levels of reality, as in any other world of experience.

Let us then agree that the statement “I write” serves the purpose of pinning down a first level of reality, which I have, explicitly or otherwise, to take account of in any operation that creates a rapport between diverse levels of reality in writing, and even between things written and things not written. This first level may be useful to me as a platform on which to erect a second level, which may even belong to a reality utterly different from the first, and indeed refer to a different stratum of experience.

For example, I might write, “I write that Ulysses listens to the song of the Sirens,” an incontrovertible statement that bridges the gap between two worlds that are not contiguous: the immediate and empirical world in which I am and am writing, and the mythical one in which it always happens that Ulysses, tied to the ship’s mast, is listening to the Sirens’ song.

The same thing might also be written, “Ulysses listens to the song of the Sirens,” leaving the “I write” understood. But if we do leave this understood we must be prepared to risk the reader’s getting confused between two levels of reality, and believing that the act of listening on the part of Ulysses takes place on the same level of reality as my act of writing this sentence.

I have said “the reader believes,” but it is clear enough that the credibility of what is written can be understood in very different ways, each one corresponding to more than one level of reality. There is nothing to prevent anyone from believing in the encounter of Ulysses with the Sirens as a historical fact, in the same way as one believes in the landing of Christopher Columbus on October 12, 1492. Or else we may believe it by feeling ourselves struck by the revelation of a truth beyond perception that is contained in the myth. But here we enter a field of religious phenomenology in which the written word would merely act as a spur to meditation. However, the credibility that interests us here is neither of these, but the kind of credibility peculiar to the literary text, in parentheses, as it were, matched on the reader’s part by an attitude Coleridge defined as “suspension of disbelief.” This suspension of disbelief is the condition on which the success of every literary invention depends, even if it is admittedly within the realm of the fabulous and incredible.

We have considered the possibility that the level of “Ulysses listens” might be put on a par with that of “I write.” But this balance between the two levels could occur only if you, the reader, believed that the statement “I write” also belonged on a literary or mythological level. The “I” that is the subject of “I write” would then become the “I” of a fictional or mythological character—such as Homer, in fact. For clarity’s sake, let us put our sentence in the following manner: “I write that Homer tells that Ulysses listens to the Sirens.” The statement “Homer tells” may be placed on a level of mythical reality, in which case we will have two levels of mythical reality, that of the fable narrated and that of the legendary blind bard inspired by the Muses. But the same statement might be placed on a level of historical or (better) philological reality. In that case, by “Homer” we mean the individual or collective author with whom scholars busy themselves over the “Homeric question,” and the level of reality would then be identical or contiguous to that of “I write.” (Notice that I have not written “Homer writes” or “Homer sings,” but “Homer tells,” so as to leave both possibilities open.)

The way in which I have formulated the sentence makes it natural to think that Homer and I are two distinct persons, but this could be a wrong impression. The phrase would be exactly the same if it had been written by Homer in person, or in any case by the real author of the Odyssey, who at the moment of writing splits into two first persons: the empirical “I” who pens the words on the page (or dictates them to a scribe) and the mythical character of the blind bard, visited by divine inspiration, with whom he identifies himself.

In the same way, nothing would change if the “I” were the “I” who is speaking to you, while Homer, of whom I speak, were also “I”: that is, if what I attribute to Homer were my own invention. This procedure would be clear at once if the phrase ran: “I write that Homer tells that Ulysses discovers that the Sirens are mute.” In this case, in order to obtain a particular literary effect, I apocryphally attribute to Homer my own inversion, or distortion, or interpretation of the Homeric narrative. (In fact, the idea of the silent Sirens is Kafka’s, and we must realize that the “I” who is the subject of the sentence is Kafka.) But even without turning things upside down, the countless authors who in recasting an earlier author have rewritten or interpreted a mythical (or at least a traditional) tale have done this to communicate something new, while still remaining faithful to the image of the original; and with all of them, in the “I” of the writing first person one can distinguish one or more levels of mythical or epic reality that draw material from the collective imagination.

Let us go back to the sentence we started with. Every reader of the Odyssey knows that more exactly it ought to be written, “I write that Homer tells that Ulysses says: I have listened to the song of the Sirens.”

In the Odyssey, in fact, the adventures of Ulysses in the third person surround and contain other adventures of Ulysses in the first person, narrated by him to Alcinoüs, king of the Phaeacians. If we compare one with the other, we find that the difference between them is not simply grammatical. The adventures told in the third person have a psychological and emotional dimension that the others lack, and in them the supernatural presence consists in appearances of the Olympian gods in the guise of ordinary mortals. On the other hand, the adventures of Ulysses in the first person belong to a more primitive repertoire of myth, in which ordinary mortals and supernatural beings meet face to face; a world peopled by monsters, Cyclopses, Sirens, enchantresses who change men into pigs, and in fact the whole pre-Olympian pagan world of the supernatural. We may therefore define these as two different levels of mythical reality, to which there are two corresponding geographical realities. One corresponds to the historical knowledge of the time (the voyages of Telemachus and the homecoming to Ithaca), while the other belongs to fable and results from a juxtaposition of the most heterogeneous traditions (the travels of Ulysses as told by Ulysses himself). We may add that between the two levels is the island of the Phaeacians, which was the ideal place that gave birth to the narrative—a utopia of human perfection, outside the bounds of history and geography.

I have dwelt on this point because it serves to show how the different levels of reality may be matched by different levels of credibility—or, to put it better, a different suspension of disbelief. Assuming that a reader “believes” in the adventures of Ulysses as told by Homer, this same reader might judge Ulysses to be a mere braggart in all that Homer makes him say in the first person. But let us be careful not to confuse levels of reality (within the work) with levels of truth (referring to things outside it). For this reason we ought always to bear in mind the entire sentence: “I write that Homer tells that Ulysses says: I have listened to the song of the Sirens.”

This is the formula that I put forward as the most complete, and at the same time the most compact, model of the connecting links between levels of reality in works of literature. Every part of this sentence may be linked to various sets of problems. I shall now give a number of suggestions, starting again from the beginning.

I write

This statement, “I write” (or “I am writing”), is connected with the whole field of problems—particularly fertile in this century—concerning what has been called metaliterature, and the analogous problems of metatheatre, metapainting, and so on. We already mentioned the play within the play while speaking of Shakespeare; there is no dearth of examples in the work of other playwrights, from Corneille’s Illusion comique to Pirandello’s Six Characters in Search of an Author. But in the last few decades these metatheatrical and metaliterary processes have acquired fresh importance, with foundations of a moral or epistemological nature, in opposition to the illusoriness of art, to the claim made by realism to lead the reader or spectator to forget that what he has before his eyes is an operation performed by means of language, a fiction worked out with an eye toward a strategy of effects.

The moral and indeed pedagogical motive is dominant in Brecht, with his theory of the epic theatre of alienation: the spectator must not abandon himself passively and emotionally to the illusion on the stage, but must be urged to think and to participate.

A process of theorization based on structural linguistics, on the other hand, forms the background to research carried out in French literature during the past fifteen years, and in both critical thought and creative practice this puts the material side of writing—the text itself—firmly in the foreground. We need only mention the name of Roland Barthes.

I write that Homer tells

Here we enter a very extensive field, the splitting or multiplication of the subject of the verb “to write”; and it is a field in which any really exhaustive theoretical work has yet to be done.

We might begin with a habit characteristic of the writers of chivalric romances, that of claiming a hypothetical manuscript as a source. Even Ariosto pretends to refer back to the authority of Turpin. And Cervantes introduces the figure of an Arab writer, Cid Hamet Benengeli, between himself and Don Quixote. Not only that, but Cervantes supposes a kind of contemporaneousness between the action narrated and the writing of the Arabic manuscript, so that Don Quixote and Sancho Panza are aware that the adventures they are having are those written by Benengeli, and not the ones written about by Avellaneda in his apocryphal second part of Don Quixote.

A still simpler method is that of suggesting that the book is written in the first person by the protagonist. The first novel that may be considered entirely modern was not published under the name of the author, Daniel Defoe, but as the memoirs of an obscure sailor from York, a certain Robinson Crusoe.

All this brings me little by little to the heart of the problem: the successive layers of subjectivity and feigning that we can discern underneath the author’s name, and the various “I”s that go to make up the “I” who is writing. The preliminary condition of any work of literature is that the person who is writing has to invent that first character, who is the author of the work. That a person puts his whole self into the work he is writing is something we often hear said, but it is never true. It is always only a projection of himself that an author calls into play while he is writing; it may be a projection of a real part of himself or the projection of a fictitious “I”—a mask, in short. Writing always presupposes the selection of a psychological attitude, a rapport with the world, a tone of voice, a homogeneous set of linguistic tools, the data of experience and the phantoms of the imagination—in a word, a style. The author is an author insofar as he enters into a role the way an actor does and identifies himself with that projection of himself at the moment of writing.

The author-cum-character is both something less and something more than the “I” of the individual as an empirical subject. He is something less because (for example) Gustave Flaubert the author of Madame Bovary excludes the language and visions of Gustave Flaubert the author of La Tentation de Saint Antoine or Salammbô. He rigorously cuts down his inner world to the set of data that make up the world of Madame Bovary. And he is something more because the Gustave Flaubert who exists only in relation to the manuscript of Madame Bovary partakes of a far more compact and well-defined state of being than does the Gustave Flaubert who, while writing Madame Bovary, knows that he was the author of La Tentation and is about to be the author of Salammbô. He knows that he is continually oscillating between one world and another, and that in the end all these worlds flow together and unify in his mind.

The example of Flaubert may be used to verify the formula I have suggested if we translate this into a series of projections or lantern slides. Gustave Flaubert the author of the complete works of Gustave Flaubert projects outside himself the Gustave Flaubert who is the author of Madame Bovary, who in turn projects from himself the character of a middle-class married woman in Rouen, Emma Bovary, who projects from herself that Emma Bovary whom she dreams of being.


Each element projected reacts in its turn on the element that projects it; it transforms and conditions it. So the arrows go not in one direction only, but in both:


All we then have to do is connect the last term with the first—that is, establish the circular movement of these projections. It was Flaubert himself who gave us a precise clue to this with his famous phrase “Madame Bovary, e’est moi.”


How much of the “I” who shapes the characters is in fact an “I” who has been shaped by the characters? The further we go toward distinguishing the various levels that go to make up the “I” of the author, the more we realize that many of these levels do not belong to the author as an individual but to collective culture, to the historical period or the deep sedimentary layers of the species. The starting point of the chain, the real primary subject of the verb “to write,” seems ever more distant from us, more rarefied and indistinct. Perhaps it is a phantom “I,” an empty space, an absence.

In order to acquire more solid substance, the “I” can become a character, and indeed the protagonist of the written work. But we need only think of the extremely subtle pages that Gianfranco Contini devotes to the “I” of the Divine Comedy to realize that it, too, can be split into a number of persons, rather like the “I” who is speaking in Proust’s A la recherche du temps perdu.

With the “I” who becomes a character we are shifting our focus from “I write that Homer tells” to “Homer tells that Ulysses …

Homer tells that Ulysses

The author-protagonist brings an internal subjectivity to the written world, a figure endowed with a distinctness of his own—often a visual and iconic distinctness—which seizes the imagination of the reader and acts as a device to connect the different levels of reality, or even to bring them into being and enable them to take on form in the course of writing.

The character Don Quixote clears the way for the clash and encounter of two antithetical languages, or, rather, of two literary worlds without any ground in common: the chivalric-supematural and the picaresque-comic. It opens up a new dimension, or, rather, two: an extremely complex level of mental reality and a representation of the environment which we might call realistic, but in a completely new sense compared with picaresque “realism,” which consisted of a repertoire of stereotyped images of poverty and squalor. The sun-drenched country roads on which Don Quixote and Sancho Panza meet with friars carrying parasols, muleteers, ladies in sedan chairs, and flocks of sheep constitute a world that had never before been written about. It had never before been written about because there had never been a reason for writing about it, whereas here it fulfills a need: it is the reverse of the inner reality of Don Quixote, or, better still, the background against which Don Quixote projects his own codified interpretation of the world.

Don Quixote is a character endowed with an unmistakable iconic quality and inexhaustible inner riches. But this does not mean that to perform the function of protagonist a character necessarily has to have such depth. The function of the character may be likened to that of an “operator,” in the mathematical sense of the word. As long as his function is well defined, he can be merely a name, a profile, a hieroglyphic, a sign.

When we read Gulliver’s Travels we know very little about Dr. Lemuel Gulliver, ship’s doctor in the Royal Navy. His substance as a character is infinitely scantier than that of Don Quixote, yet his is the personality that we follow throughout the book, and the one that brings it into being. This is because, even though it is hard to define Lemuel Gulliver psychologically or facially, his function as an “operator” is very clear, in the first place as a giant in a world of dwarfs and then as a dwarf in a world of giants. And this operation in terms of size is the easier interpretation, whereby Gulliver “works” as a character even for children who read Swift’s book in simplified texts. But the real operation which he manifests (and here I am thinking of a very persuasive essay on this subject by an Italian scholar, Giuseppe Sertoli, published this year) is that of the contrast between the world of logico-mathematical reason and the world of bodies, of their physiological materiality and various cognitive experiences and various ethical and theological concepts.

Ulysses says:

Punctuated with a colon. This colon is a very important articulated “joint,” and I would call it the headstone of narrative at all times in all lands. Not only because one of the most widespread structures of written narrative has always been that of stories inserted into another story that acts as a frame, but also because where the frame does not exist we may infer an invisible colon that starts off the discourse and introduces the whole work.

I will confine myself to touching on the salient problem. In the West, the novel was born in Hellenistic Greece, taking the form of a main narrative into which secondary narratives, told by the various characters, were inserted. This method is characteristic of ancient Indian narrative, where, however, the structure of the story in relation to the point of view of the narrator obeys rules far more complicated than in the West. (I am here relying on a monograph published in 1914, Sur l’origine indienne du roman grec, by F. Lacôte, an expert on Indian literature.) Also based on Indian models are those collections of stories inserted into a narrative that acts as a frame, both in the Islamic world and in medieval and Renaissance Europe.

All of us have in mind The Arabian Nights, in which all the stories are contained within the general framework of the tale of the Persian king, Shahryar, who kills his brides after their wedding night, and of his bride Scheherazade, who succeeds in postponing this fate by telling wonderful stories and stopping just at the climax. Besides the tales told by Scheherazade, there are others narrated by characters in her tales. In other words, the stories are like boxes within boxes—as many as five of them at a time. Here I am relying on an essay called Les Hommes-récits by Tzvetan Todorov, who has studied the enchâssement of the tales in The Arabian Nights and in Potocki’s Manuscrit trouvé à Saragosse (Poétique de la prose [Paris: Seuil, 1971]).

Borges speaks of one of the Arabian Nights, number 602, the most magical of all, in which Scheherazade tells Shahryar a tale in which Scheherazade tells Shahryar a tale, etc., etc. In the translations of The Arabian Nights that I have at hand, I have never been able to find this 602nd Night. But even if Borges invented it he did well, because it represents the natural culmination of the enchâssement of the tales.

From the point of view of levels of reality, I should say that the enchâssement of The Arabian Nights does indeed create a structure in perspective, but to my reading, at least in the only way we can read them, these tales appear to be all on the same plane. In them we can distinguish two very different types of narrative: the magical type of Indian and Persian origin, with its genies, flying horses, and metamorphoses, and the Arab-Islamic storytelling type of the Baghdad cycle, with the caliph Harun-al-Rashid and Jafar the vizier. But the tales of both types are put on the same plane, both structurally and stylistically, and as we read we slide from one type to the other, as the eye slides over the surface of a tapestry.

In the prototype of literary storytelling in the West, Giovanni Boccaccio’s Decameron, between the tales and the framework there is a clean stylistic split that highlights the distance between the two planes. The framework of the ten days of the Decameron describes the pleasant life lived in their country abode by the seven women and three men of this happy band of storytellers. We are on a plane of stylized reality, uniformly pleasant, refined in a mannered fashion, without contrasts, without characterization; nothing but descriptions of the weather and the landscape, of the pastimes and conversations of this playful little court that every day elects a queen and ends the day with a poem. The tales narrated, however, constitute a catalogue of the narrative possibilities becoming available to language and culture at a time when the variety of living forms was a new thing, coming into its own at that very moment. Each novella reveals an intensity of writing and of representation that radiates outward in all sorts of different directions in such a way as to stress these directions in comparison with the frame. Does this mean that the frame is a merely decorative element? If we said this, we would be forgetting that the framework of the tales, that earthly paradise of the elegant court, is itself contained within another frame, tragic, deathly, hellish: the plague in Florence in 1348, described in the introduction to the Decameron. It is the stark reality of a world at the end of the world, the plague as a biological and social catastrophe, that makes sense of the utopia of an idyllic society governed by beauty, gentleness, and intelligence. The chief product of this utopian society is the short story, and the story reproduces the variety and nervous intensity of a world that is lost, the laughter and tears by now erased by Death the Leveler.

Let us now see what is inside the frame.

I have listened to the Sirens’ song

I could also have said: I have blinded Polyphemus the Cyclops, or: I have eluded the enchantments of Circe. But if I have chosen the episode of the Sirens it is because this enables me to introduce a further transition in perspective with the narrative of Ulysses, a further level of reality, contained in the Sirens’ song.

What do the Sirens sing? One possible hypothesis is that their song is nothing more or less than the Odyssey. The tendency of the poem to incorporate itself, to reflect itself as in a mirror, appears a number of times in the Odyssey, especially at banquets when the bards sing. And who better than the Sirens could endow their own song with this function of magic looking glass?

In this case we would have a case of the literary process that André Gide defined by the heraldic term mise en abyme. This occurs when a work of literature includes another work resembling the first—that is, when a part of it reproduces the whole. We have already mentioned the play within the play in Hamlet and the 602nd Night according to Borges. Examples also extend to painting, as in the mirror effects in the pictures of van Eyck. I will not dwell on the mise en abyme, because I need only refer you to an exhaustive study published a few months ago: Lucien Dällenbach, Le Récit spéculaire (Paris: Seuil, 1977).

But what the text of the Odyssey tells us about the Sirens’ song is that the Sirens say they are singing and wish to be heard, and that their song is the best that can be sung. The final experience taken into account in the narrative of Ulysses is a lyrical and musical experience on the borders of the ineffable. One of the finest passages in Maurice Blanchot interprets the Sirens’ song as a “beyond” of expressive possibilities from which Ulysses, having experienced its ineffability, withdraws, falling back from the song itself to his account of the song.

In order to verify my formula I have so far made use of narrative examples, choosing from the classics in verse or prose or in the theatre, but always with a story to tell. Now that I have got to the Sirens’ song, I ought to go back over my entire argument to see if it can be applied, as I think it can, point by point to lyric poetry, and shed light on the various levels of reality that the “operation” of poetry traverses. I am convinced that this formula, with very small modifications, can be transcribed with Mallarmé in Homer’s place. Such a formulation might perhaps enable us to pursue the Sirens’ song, the ultimate point writing can attain, the final core of the written word, and perhaps in the wake of Mallarmé we would arrive at the blank page, at silence, at absence.

The path that we have followed—the levels of reality evoked by literature, the whole gamut of veils and shields—may perhaps stray off into infinity, may perhaps encounter nothingness. As we have witnessed the disappearance of the “I,” the primary subject of the verb “to write,” so the ultimate object eludes us. Perhaps it is in the field of tension between one vacuum and another that literature multiplies the depths of a reality that is inexhaustible in forms and meanings.

Now, right at the end of my talk, it occurs to me that I have been speaking throughout of “levels of reality,” whereas the topic of this conference reads “The Levels of Reality.” Perhaps the fundamental point of my talk is exactly this: literature does not recognize Reality as such, but only levels. Whether there is such a thing as Reality, of which the various levels are only partial aspects, or whether there are only the levels, is something that literature cannot decide. Literature recognizes the reality of the levels, and this is a reality (or “Reality”) that it knows all the better, perhaps, for not having come to understand it by other cognitive processes. And that is already a great deal.