by John Barth
Perhaps this is the place to review some elementary propositions about reality and fantasy, which I’m sure have been discussed in a sophisticated way in various sessions of this conference. Ludwig Wittgenstein, in the Tractatus Logico-Philosophicus, defines the world (which is to say, reality) as being “everything that is the case.” The Cabalists, whom writers as different as I. B. Singer and Jorge Luis Borges have found to be a rich source of literary metaphor, maintain that this reality, our reality, is God’s text, his significant fiction—I believe Mr. Singer might even say God’s executed fantasy. Arthur Schopenhauer, whose importance to Borges I’ll come back to, goes farther and declares that our reality, whether or not it’s God’s fiction, is our representation, as it were our fiction: that relations, categories, concepts such as differentiation, time and space, being and not-being—all are ours, not seamless nature’s. Eastern philosophy teases these paradoxes out; their bottom line is that capital-R Reality—to a greater or lesser extent, but strictly speaking—is our shared fantasy.
I trust that this conference will agree, in the main, that the difference between the fantasy we call reality and the fantasies we call fantasy has to do with cultural consensus and with one’s manner of relating to the concept-structure involved: What we call the real world, we relate to as if it were the case. Psychopathological fantasies are more or less individual concept-structures markedly at variance with the cultural consensus and related to as if they were the case: If you conceive yourself to be Napoleon and act upon your conviction, the rest of us will put you away. “Normal” fantasies are more or less individual concept-structures significantly at variance with the cultural consensus but not related to as if they were the case: e.g. night-dreams (from our waking point of view), daydreams, and aesthetic fictions both “realistic” and “fantastic.” On this ontological level, all fiction is fantasy. (The electronic-computer microworlds dealt with by investigators in the field of artificial intelligence would seem to belong to some third category: perhaps secondary fiction in a sense different from Gardner’s. The AI computer constructs a world along parameters laid down by its programmer—a microworld often at considerable variance indeed with the programmer’s cultural consensus—and it relates to that microworld as if it were real; but we don’t call either the computer or the programmer psychopathological—at least not ipso facto. I confess to not having thought this aspect of the matter through very carefully: I’m a storyteller, not a philosopher.)
Aesthetic realism, then, is any set of artistic conventions felt by people on a particular level of a particular culture at a particular period to be literally imitative of their imagination of the actual world. It goes without saying that one generation’s or culture’s realism is another’s patent artifice—witness for example the history of what has passed for realistic dialogue and characterization in Hollywood movies from Humphrey Bogart to Robert de Niro. It likewise goes without saying that what the inexperienced find realistically convincing, the experienced may not, and vice-versa: The birds peck at Apelles’s painted grapes (almost the only thing we’re taught about classical Greek painting); the innocent frontiersman rises from his seat at a nineteenth-century showboat melodrama to warn the heroine against the villain’s blandishments. On the other hand, zoo zebras ignore a life-size color photograph of a zebra—they don’t know what it represents—and the Colombian novelist Gabriel García Márquez tells us that what we gringos take for surrealism in his fiction is everyday reality where he comes from.
As for aesthetic irrealism —fantasy of the sort addressed by international conferences on the fantastic—it must consist of any set of artistic principles and devices, conventional or otherwise, felt by people on a particular level of a particular culture at a particular time to be enjoyable and/or significant though understood to be not literally imitative of their imagination of the actual world. Consider the ghost of Hamlet’s father. For most of us and for many Elizabethans, that ghost in Shakespeare’s play is/was a device of fantastic literature. For many other Elizabethans and some of us—those who believe in ghosts—it was/is a device of realistic literature. We don’t know which it was for the playwright or for particular players of the role. For the character Hamlet, the ghost is no device of any sort; it is reality, as it may well have been for some innocent rube scared out of his skin in the stalls of the Globe. Hamlet’s mother Gertrude calls the ghost Hamlet’s fantasy (“… the very coinage of your brain”: III, iv), even his psychopathological fantasy (“Alas, he’s mad”: III, iv): She neither sees nor hears the ghost when Hamlet does, in her presence. But there’s the complication that not only the watchmen but also the antisupernaturalist university student Horatio all see what Hamlet sees but apparently don’t hear what he hears (this after the guard Marcellus had complained, “Horatio says ’tis but our fantasy”: I, i)—etc., etc. We are in the paradoxical world of realistic fantasies like Kafka’s and fantastic realities like Edgar Poe’s version of Scheherazade’s last tale: the one that King Shahryar refuses to swallow, having to do with steamboats and railways and other preposterosities. I shall add that I myself find the fantastic device of Hamlet’s father’s ghost a good deal more believable than the realistic device of the accidental exchange of poisoned swords in midst of Hamlet’s duel with Laertes in Act V. So it goes—and it is time to wander back to our subject.
I remarked that the devices of aesthetic fantasy may be conventional or otherwise. In the opinion of Jorge Luis Borges, the most ubiquitous devices of fantastic literature are four: the double, the voyage in time, the contamination of reality by irreality, and the text within the text.
Of this last, my topic, we’re all familiar with such classic examples as The 1001 Nights, in which Scheherazade entertains King Shahryar with stories in which the characters sometimes tell one another stories in which (in a few cases) the characters tell further stories. Closer inspection reveals that the real “frame” of The 1001 Nights is not the relation between Scheherazade and Shahryar, but something “farther out” and more ancient: the relation between the reader, or listener, and the unspecified teller of the story of Scheherazade. I mean this literally: The opening words of the tale (after the invocation to Allah) are “There is a book called The 1001 Nights, in which it is said that once upon a time…” etc. In other words, The 1001 Nights is not immediately about Scheherazade and her stories; it is about a book called The 1001 Nights, which is about Scheherazade and her stories. That book is not the book we hold in our hands, with wonderful notes by Richard Burton; nor is it quite the book that Shahryar on the 1002nd morning orders to be written. I don’t know what, exactly, that book is, or where. I have asked my friend William H. Gass—a professional philosopher as well as a professional storyteller—please to locate that book for me; he’s not sure where it is either. To think about that book very long is to invite vertigo.
The classical invocation to the muse of Greek and Roman epic literature can be regarded similarly as a radical framing device. The “outside” or ground-story of the Odyssey may be said to be, not the situation of Odysseus striving home from Troy, but the situation of the bard who in the opening lines sings (in Albert Cook’s translation) “Sing, Muse, of that man of many turns, the wanderer,” etc.; “begin anywhere you like.” To which the Muse in effect replies: “Since you ask, the story goes this way: All the other Greeks had got home long since,” etc. My experience and intuitions both as a professional storyteller and as an amateur of frame-tale literature lead me to suspect that if the first story ever told began “Once upon a time,” the second story ever told began “Once upon a time there was a story that began ‘Once upon a time.’ ” Furthermore—since storytelling appears to be as human a phenomenon as language itself—I’d bet that that second story was no less about “life” than the first.
But let me tell you the story of my romance with this second sort of stories: tales within tales. You’ve heard its beginning: that student, once upon a time, pushing his book-cart through the stacks of the
Johns Hopkins classics library and surreptitiously reading the fantastic literature he was supposed to be filing: The 1001 Nights, The Ocean of Story, the Panchatantra, the Metamorphoses, the Decameron, Pentameron, Heptameron, and the rest. A good many years later, that student found himself metamorphosed into a storyteller as well as a story-reader, and a professor at The State University of New York at Buffalo to boot, one of whose perquisites (this was the palmy mid-1960s) was a graduate-student research assistant. How I could have used one earlier, at Penn State, when I was writing The Sot-Weed Factor! How I could have used one later, at Johns Hopkins, when I was writing the novel LETTERS! But as it happened, my fiction just then in progress required no particular research.
To keep the situation honest, I therefore resolved to implement my old affection for frametale literature with a reasonably thorough, if non-professional, investigation of the genre—of which, in fact, there seemed to be little in the way of general examination beyond some useful checklists in German‡ and some side-glances by Chaucerians.§ My objective was neither to publish an essay on the subject nor to teach a course in it: simply to ask certain questions of the existing corpus of such literature in order to satisfy a long-standing curiosity and, perhaps, to discover something about that ancient narrative convention which might inspire a story of my own: a story which, whatever else it was about, would also be about stories within stories within stories.
I shall digress again briefly here to remark that it does not appear to matter to the Muses whether a writer invokes them out of a heroic wish like Alexander Solzhenitsyn’s, to expose and destroy an oppressive system of government, or a decadent wish like Flaubert’s, to write a novel “about nothing”: Their decision to sing or not to sing seems based on other considerations. The Muses are a less responsible committee, in the moral sphere, than is the Swedish Academy.
Well. We proceeded to interrogate literature, fantastic and otherwise: I mean that research assistant, the Muses, and me. Here, briefly, are some of the findings of our leisurely and gentle interrogation and some of the things that those findings suggested to me.
First, as might be expected, we found it necessary in designing our questionnaire to distinguish categories of frametale literature. Our first distinction was between incidental or casual frames and more or less systematic frames: It is in fact more of a spectrum or continuum than a distinction. In the first category we put such unforgettable but incidental stories-within-stories as Pilar’s story of the killing of the fascists in Chapter 10 of Hemingway’s For Whom the Bell Tolls, or Ivan’s tale of the Grand Inquisitor in Book V, Chapter V of Dostoevsky’s The Brothers Karamazov; also the incidental romances with which Cervantes interrupts the adventures of Don Quixote; and, for that matter, such classical retrospective expositions as Odysseus’s rehearsal to the Phaeacians of his story thus far (Books IX-XII of the Odyssey) and Aeneas’s ditto to Dido (Books II and III of the Aeneid). More fiction than not, I suppose, frames some incidental anecdote or delayed anecdotal exposition. We decided to confine our attention to the other end of the spectrum: stories that programmatically frame other stories.
There we soon found ourselves making further taxonomical distinctions, which I shall merely illustrate. There is for example what I think of as the Dante/Chaucer continuum: on the one hand, stories Like The Divine Comedy, in which the frame (Dante’s impasse in the Dark Wood and his detour through Hell, Purgatory, and Heaven) is at least as conspicuous and as dramatically developed as the stories told along the way, most of which in Dante’s case are plotless moral exempla or extended epitaphs; on the other hand, stories like Chaucer’s Canterbury Tales or Boccaccio’s Decameron, in which the framed stories are dramatically complete, but the frame-story—the pilgrimage to and from Canterbury; the retreat of ten young Florentine ladies and gentlemen from the plague of 1348—is vestigial, rudimentary, incomplete, or dramaturgically static. I wish now that we had also kept tab of comparatively realistic frames for comparatively fantastic stories—pretty much the case with the Odyssey and the Nights, for example—and vice-versa. But we didn’t.
Next, it seemed useful to distinguish between frame-stories with a single frame—such as Dante’s, Chaucer’s, and Boccaccio’s, and in fact almost all frametale literature at the first level of its framing—and the very much rarer cases of serial primary frames: as if for example the pilgrimage to Canterbury were but the first in a series of linked Chaucerian ground-narratives whose characters proceed to tell their several stories. Two conspicuous examples of this rarer species are the eleventh-century Sanskrit Ocean of Story, a mammoth work of great structural complexity involving a series of very intricate primary frames,|| and Ovid’s Metamorphoses, whose armature is an extraordinarily subtle and graceful series of linked primary frametales.
Further, we saw fit to ask of each of the several hundred specimens of frametales and quasi-frametales that we buttonholed whether it comprises merely two degrees of narrative development—tales within a tale, such as Dante’s and Boccaccio’s—or three or more such degrees: tales within tales within a tale, etc. In the oriental literature, we found, it is not uncommon at all for the characters in a second-degree story to tell stories of their own. Where this movement to the third degree occurs more than once—e.g., in The 1001 Nights —the second degree of narrative (Scheherazade’s stories) becomes a serial frame within a single frame (the story of Scheherazade). Where the characters on the third level of narrative involvement more than once tell further stories, as in the Panchatantra, we have stories serially framed within serial frames within a single frame. The Panchatantra in fact moves to as many as five degrees of narrative involvement, as does The Ocean of Story —whose primary frame, we remember, is itself serial. Indeed, The Ocean of Story manages to engulf the whole Panchatantra as one of its serial frames, and the Vetalapanchavimsati (25 Tales by a Vampire) as another.
Such oriental complexity is uncommon in occidental literature, if we ignore that sort of quasi-framing which I spoke of earlier: invocations to the muse and formulae like “There is a story about a man who” etc. But there are pleasant exceptions: Ovid shifts his Metamorphoses to at least four degrees of narrative involvement—e.g., in Book VI, where in the course of Ovid’s ongoing story the Muse tells Minerva the story of the contest between the Muses and the daughters of Pierus, in course of which story the muse Calliope tells the jury of nymphs the story of the rape of Proserpine, in course of which story Arethusa tells Ceres the story of her own rape: a tale within a tale within a tale within a tale. It is managed so subtly that it goes almost unnoticed unless one happens to be reading a translation with conventional English punctuation and sees such strange accumulations of double and single quotation-marks as ‘ “ ‘ “. The Saragassa Manuscript, Tzvetan Todorov points out, reaches no fewer than five degrees of narrative involvement: a veritable Sargasso Sea of Story. Even more remarkable, though vestigial, is the frame of Plato’s Symposium. We all remember that the guests at Agathon’s banquet take turns making speeches or telling stories about love, and that the climactic speech is Socrates’s famous description of the Ladder of Love, whose final rung he says was explained to him by a lady named Diotima. Many of us will have noticed or had it pointed out to us that the story of Agathon’s banquet itself is not told us directly by Plato, but by a fellow named Apollodorus, who is telling it to an unnamed friend. In fact, Plato writes that Apollodorus reports that he has the story from a disciple of Socrates’s named Aristodemus, who was among Agathon’s party guests. We are given the conversation between Apollodorus and this Aristodemus. However, Apollodorus is not telling his unnamed friend directly what Aristodemus told him: Apollodorus tells his unnamed friend the story of his being importuned in the street two days earlier by yet another friend, Glaucon, who wanted to hear what Aristodemus had reported to Apollodorus of Socrates’s speech at Agathon’s party. When we sort it all out, we discover that:
1. Apollodorus is telling his unnamed friend
2. the story of Apollodor
us’s telling Glaucon
3. the story of Aristodemus’s telling Apollodorus
4. the story of Socrates’s telling Agathon’s company
5. the story of Diotima’s telling Socrates
6. the story of the Topmost Rung on the Ladder of Love.
That is, we are about as many removes from Diotima’s story as there are rungs on the Ladder of Love itself, even before we add the next frame out—Plato’s telling all this to the reader—and the next frame out from there: my reminding you what Plato tells the reader. If, as I hope, some of you tell your lovers tonight the story of these remarks of mine, you will be involved in a frametale of nine degrees of narrative complexity, unapproached in the actual corpus of frametale literature.
To return to that corpus: Its formal possibilities can be visualized, and actual specimens schematized, by conventional outline format, if we ignore the rule of logical outlining that forbids a I, say, unless there’s a II, an A without a B, etc. We are dealing here not with logic but with spellbinding. The Canterbury Tales or the Decameron we might begin to schematize as in Figure 1—
—where I represents the pilgrimage from Canterbury or the Florentine aristocrats’ retreat from the plague, and the upper-case letters represent the several tale-tellers and their tales. Where a character tells two or more tales serially, we might improve the notation as in Figure 2:
