Homer's Forge

On the relationship between real and fictional worlds

Dale Peck

Artwork by Yohei Oishi

Ours is "the best of all possible worlds," Gottfried Leibniz famously told us in 1710, in what was either the last gasp of the ancien régime or the first hesitant gurgle of Enlightenment. Fifty years later, Voltaire put the nail in the coffin of Leibniz's argument with a retort that became, if not more famous than the original, then inseparable from it: "Then what must the others be like?" Voltaire's riposte sidestepped the crux of Leibniz's thesis, which was that God had created a world whose manmade evils were a byproduct of that most precious and problematic of His gifts, free will. Voltaire wasn't an atheist, but he wasn't much concerned with God's plan either, at least when it came to public policy; which is to say, iniquity didn't interest him as much as inequity. Like many well-meaning Enlightenment thinkers, Voltaire sought a temporal mode of being that was most just for its citizens, or at any rate least unjust. Not utilitarian per se, but one that made an honest attempt to reconcile material well-being with the more ephemeral problem of happiness. This state of affairs is easier to define in the negative than the positive—at least in the absence of a deistic outlook like Leibniz's—which is why dystopic or satiric tracts such as Candide have always outnumbered idealistic treatises along the lines of Leibniz's Théodicée or Thomas More's Utopia, whose form Voltaire borrowed to shape Candide. Leibniz understood this as well as Voltaire, and, while his pragmatism has more than a hint of apology, it also reflects a tolerance of and sensitivity to human nature that the hair shirt–wearing More certainly didn't possess, or couldn't acknowledge. Like More, Voltaire couldn't accommodate himself to the world as it was, but unlike More he couldn't come up with anything better. He opted instead for irony, aphorism, le mot juste, leaving the specter of all those other possible worlds to readers' imaginations.

While it's tempting to claim that writers have been trying to answer Voltaire's question ever since he posed it, the truth is they started doing it centuries—millennia—before Candide. Hence Utopia, hence the Republic and the Garden of Eden and every other creation myth in which the world starts out a paradise only to be debased by a sinful or flawed humanity. The world that Homer gave us in The Iliad and The Odyssey, that Ovid created in The Metamorphoses, Dante in the Comedia, Cervantes in Don Quixote, on down to the latest pulpy thrillers and romances on the bestseller list, all are iterations of Voltaire's "others": worlds in which the varieties of injustice and unhappiness are as manifold as their characters. Yet at the same time they are also manifestations of Leibniz's "best of all possible worlds," for, just as God promises eternal reward to those who choose to observe his laws, so too do fictional universes force meaning on life's chaos—redemption, significance, the simple consolation of order. Some readers open a novel looking for Voltaire's infinite varieties of suffering, others for Leibniz's mitigated perfection, but in either case the question remains: what is the relationship between these invented worlds and our own? By which I mean the real world, the one readers seek to escape when they plunge themselves into an author's artificial universe.

In fact, my interest in this question arose under far more pedestrian auspices than this introduction might suggest. In 1999, I began teaching in the New School's Graduate Writing Program, and over the course of the next couple of years I noticed that the majority of my students took setting—i.e., the world, and its fictional counterpart—for granted. A remarkably high percentage wrote either about their hometowns, typically from a child's point of view, or New York City, typically from the point of view of a young person newly arrived here. In response, I decided to teach a series of books that I thought demonstrated setting to its full advantage. Though little more than a kneejerk solution to a localized problem, it turned out to be far more rewarding than I could have expected. Halfway through that first semester, I came up with a working thesis that became the starting point for the course every time I taught it thereafter, a concept that has come to shape the way I think about fictional universes, both in others' work and my own: namely, that setting, far from being a distant number three in the great triumvirate of character, narrative, and setting, is in fact one half of an exclusive dyad—setting and character—of which narrative is a mere byproduct. A valued byproduct to be sure, akin to the meal that results from a farmer's patient husbandry of the soil, but even so, derivative rather than starting point, effect rather than cause.

On the face of it, this is a self-evident assertion. Nothing can happen in a void—a genuine void, rather than a metaphoric facsimile—and even if, by dint of some postmodern or science-fictional conceit, you place your characters in a void, the void itself becomes the setting. There must be a world for your characters to inhabit and interact with in order for there to be actions. This is a lesson filmmakers understand better than novelists. The camera doesn't discriminate between people and things, so filmmakers have no choice but to reckon with the physical universe in the same way they do their actors. Writers, however, can do as much—or, often, as little—with the world as they want. I learned this when my second novel was adapted into a film: by far the most fruitful (and humbling) part of the experience was the relocation of various scenes to more visually interesting locales, such that, a decade after the movie was made and thirteen years after I finished the book, there are several scenes that I remember taking place in the locations used in the movie, not only because they're better to look at than the ones I came up with, but also more thematically resonant.

Don't misunderstand me: a scene that takes place in a typical suburban living room can, without resorting to overly pointed dialogue (which Gore Vidal has called the laziest form of storytelling), reference the war in Iraq or the destruction of Tenochtitlan or the search for an HIV vaccine, but chances are it will involve some kind of hand tipping on the part of the writer—a crack in the plaster reminiscent of the course of the Tigris and Euphrates, say, or a photograph of an Aztec artifact on the cover of a magazine. Sometimes, of course, it's not only convenient but necessary—or, as Aristotle would put it, probable—for a scene to take place in a living room, but often it's not, and the thoughtful writer can and will transfer the same scene to a more dynamic venue, one that gives rise to thematically pointed observations, dialogue, and actions. By the same token, it should be noted that writers often place scenes in locations they fail to take advantage of. This is every bit as limiting as setting the scene in a quotidian space: if readers feel that the world around the characters is more interesting than they are, or would make them more interesting if they interacted with it, they come to regard that world as nothing more than an inexpertly painted backdrop. In fact, this could be said to be the problem with my students' choice of New York as a setting: the city deserves more than yet another naif's jaw-dropped gawking.

To illustrate this relationship, I came up with a pretty straightforward Venn diagram:


One of the reasons I find this illustration helpful is because it reminds me that a good portion of setting and character—in fact, most of both—exist solely for their own sake. If too much of setting and character overlap, the resulting narrative will appear contrived. Part of a novel's characters' significance stems from the fact that the writer chose them out of everyone else in the world; and part of what makes the things in their world compelling is the fact that there is so much else they could have interacted with—but if they had, it would have produced a different story.


There are other implications of this relationship, but before we get too deeply into them it behooves us to float some definitions. Few fields play faster and looser with its operative terms than literary aesthetics. Look at the introduction to James Wood's The Broken Estate, where the (deliberately?) slippery relationship between "realism," "realistic," "reality," "the real world," or, the ultimate bugbear, "the real," is enough to make you nuts. Similarly, it's virtually impossible to find two writers who see eye to eye on the meaning of, and distinctions between, the terms "story," "narrative," and "plot." "Setting" seems benign by comparison. Nearly everyone would agree that it refers to the environment in which a novel or story takes place. I use the word "environment" rather than "location" or "place" because the latter terms skew toward geography, when in fact the most important part of setting isn't place but milieu: culture to the physical world's nature, the law rather than the lay of the land. Which is not to say that the physical environment isn't important: it forms the crucial component of what John Gardner called "the fictional dream," locating a character in space and giving him not so much life as mortality, which is what makes life precious. Within that vast rubric, the writer is free to (indeed, must) emphasize those details that amplify some aspect of his or her theme in as specific and seemingly inevitable a manner as possible. In Roald Dahl's "Lamb to the Slaughter," Mary Maloney bludgeons her husband with a frozen leg of lamb she was about to cook for dinner; later, she serves the lamb to the detectives who investigate the murder. She could have stabbed Mr. Maloney with the knife she was using to cut the lamb or hit him over the head with the frying pan. Using the leg of lamb is not only more novel and disquieting, it also spins the story's theme in an unexpected and ultimately illuminating direction. We should be shocked by the idea of a wife killing her husband (even one who seems about to be abandoned in the sixth month of her pregnancy); instead, we're focused on the Ovidian transformation of the domestic environment. A leg of lamb? She killed him with a leg of lamb? Somehow the murder itself seems trivial by comparison. The ease with which Dahl shifts our attention reveals something about our attitudes towards violence within marriage: perhaps we're not so shocked that she killed him after all. Perhaps there are thousands of women like Mrs. Maloney, who lack nothing but a similarly rarified opportunity—a similarly specific instrument—to do the same.

Maloney inhabits an intensely cloistered world. With the exception of a brief trip to the grocery store, she spends all her time in her living room and kitchen, and her attention, when not focused on her husband, gravitates to the objects around her: a lamp, a chair, curtains, her knitting, the freezer with its fateful leg of lamb. These objects are rarely described, yet all are intensely important to Maloney's sense of place and self. So bounded is her world that we don't know if the Maloneys live in a house or an apartment, a village or a city, Ireland, England, the United States. This drift from proximal specificity to distant vagueness epitomizes the distinction between microsetting and macrosetting, i.e., the world immediately surrounding the characters and the world at large. Though both are equally material realms, the physicality of microsetting is directly experienced by the characters (and thus the reader), whereas macrosetting exists largely by inference, receding not just into the distance but into milieu as well. Borges's "The Immortals," for example, begins in "hundred-gated Thebes." The word "Thebes" locates us in space, but the Homeric epithet "hundred-gated" locates us in time—not a specific time, perhaps, but one that is generally ancient and myth-shrouded. We don't need to be told that Thebes is situated on the east bank of the Nile 800 kilometers south of the Mediterranean (thank you, Wikipedia), that it intermittently served as Egypt's capital and is the site of the great ruins of Luxor and Karnak; simply put, these things aren't important to the story, and including them would bog things down. Microsetting, however, must be much more clearly drawn, with specific sensory details, in order to locate a character in space and make him feel real (emphasis on feel, not real). Borges again: "I found that my hands were bound behind my back and I was lying in an oblong stone niche no bigger than a common grave." Note the juxtaposition of "oblong stone niche no bigger than a common grave" with "my hands were bound behind my back and I was lying...." The former is perfectly clear all by itself, but, by contextualizing it with a vivid sense of the protagonist's physical being, it gains psychological and symbolic weight, thus making a much stronger impression. A coffin looks like a really big box, until you're lying in one.

Both "Lamb to the Slaughter" and "The Immortals" point out that the division between macro- and microsetting is psychological as much as proximate—that the physical world, far from being the mundane or craft-oriented aspect of setting, demands a hefty existential toll. Take Stephen Dedalus's famous list from the beginning of A Portrait of the Artist as a Young Man:

Class of ElementsClongowes Wood CollegeSallinsCounty KildareIrelandEurope The WorldThe Universe

Macrosetting starts at the bottom, obviously, with The Universe, The World, etc., and continues through Europe and Ireland. Microsetting starts at the top, with Class of Elements, Clongowes Wood College, Sallins. But where does one stop and the other begin? In the schoolroom? The town? The county? It's impossible to pinpoint the division exactly; rather, Stephen's list is a series of concentric circles in which the familiar gradually, amorphously gives way to the strange, the foreign, the unknown, the unknowable. (In fact, the first item on the list is, of course, "Stephen Dedalus," which raises the possibility that not only is there no prima facie element of narrative, but no prima facie element of character either; there's just the world, of which human beings are simply another aspect, along with cities and mountains and suchlike.) If one of fiction's functions is to increase the sphere of what the characters know about the world, and thus what the reader knows about the world, then a well-rendered depiction of the fluid boundaries between micro- and macrosetting—between the world defined in part or in whole by what we know about it, as opposed to the world that knows and cares nothing about us—can embed that process within the landscape of the story itself, a line of communication every bit as potent as a fiber optic cable invisible beneath a surburban street.


While it's theoretically possible to describe the physical world without also describing milieu, it's extraordinarily difficult. Nature, within the world of the novel, is fundamentally bound by culture, even if that culture is represented solely by the narrational point of view. Readers of fiction don't care if the tree that falls in the woods makes a sound. They only want to know if it hits one of the characters on the way down.

On the other hand, it is impossible to describe milieu without also describing the physical world for the same reason I mentioned earlier: nothing can happen to nothing; phenomena require objects and context in order to exist. When we combine milieu and location, we think of the manmade or altered environment rather than a purely natural environment or an environmental phenomenon that possesses no existential meaning in and of itself. This presupposes a secular worldview—i.e., that some sort of omniscient entity did not make the world with a plan in mind, that storms are not symbols of divine displeasure and rainbows aren't there to remind us of God's covenant, and so on and so forth. This isn't to say that the natural world can't carry historic, cultural, or symbolic significance, only that this added meaning is nothing more than authorial commentary—sometimes wanted, more often not—or projection on the part of the character. Without endorsing this secular point of view, I'm going to adopt it for right now; later I'll try to problematize it a little. But first I want to clarify what I mean by milieu.

Broadly speaking, milieu refers to how people behave in the world of the story, and why they behave the way they do: all of the customs, laws, rituals, superstitions, and pre- and proscriptions that distinguish the raw from the cooked, and characters' attitudes toward same. The things they eat. The clothes they wear. The language they speak and the idioms, slang, and jargon they use. The types of jobs they do, and how much time they spend doing them, and whether and how they're remunerated. The broad differences—between, say, 12th century imperial Japan in The Tale of Genji and 1960s hippie paranoia in The Crying of Lot 49—are so obvious as to be generally uninteresting, but the fine distinctions—between, say, the idiom of poor blacks and whites in the rural south during the Great Depression—can set apart the writer who knows his stuff from the writer who's shamming, or slumming.

Many if not most aspects of milieu are so banal as not to be worth mentioning; and as with the physical world, the writer must consider what to put on the page and what to allow readers to fill in themselves. As with location, writers choose what to include based first on its ability to bring the world of the story to life, and then its relevance to his or her larger themes. For example: the presence of women in a US classroom is unremarkable, but in Afghanistan the presence of women in a classroom is so unusual as to require explanation. But sometimes emphasizing the obvious can put it in a different light. A writer might choose to mention that many American universities only went coeducational within the last hundred or seventy-five or fifty years, that as of 2007, American women still earn 80 cents on the dollar compared to men, and that a woman's right to choose whether to carry a baby to term is determined by a male-dominated legislature and judiciary, and thus make the point that the United States and Afghanistan, though physically far apart, exist on a clear continuity of gender-based discrimination that cuts across cultures and countries.

As all of this suggests, milieu has profound implications for characterization. Indeed, the relationship between milieu and character is so close that it would be easy to confuse an accurate depiction of milieu with characterization. It's common of workshop participants (and, at the other end of the process, critics) to say something like, "I don't believe a woman could do that to her child," or, "when you place a dark-skinned man wearing a yashmagh on a plane, the reader immediately thinks of terrorists." Readerly expectations of milieu govern everything a writer's characters do, and it's up to the writer to judge the cultural baggage associated with this or that person (or situation, or object). Characters must either conform to a perceived norm, or, if they diverge from it, the differences require some kind of comment within the story. And of course it is the divergences from expectation, from the norm, from conformity, that make a character memorable. We are fascinated with Huck Finn because, though he seems every bit a (white) boy of his place and time, he is also distinct from everyone around him. By contrast, we tend to regard Jim almost as an object because he seems to be too much of a black man of his place and time, which is to say, too much of a slave and not enough of an individual. Such concerns have played a major role in fictional narrative, in which one of the ever more prominent themes is that of self-actualization, almost always defined as overcoming the expectations of family or community or nationality—of milieu—and being true to oneself.

As a final note on definitions, I would add that the term "characterization" is misleading. When we say "character," we're talking about a person. But when we say "characterization" we're talking about what people do, i.e., actions, i.e., narrative. And so we've circled back to our original thesis: that narrative is the byproduct of setting and character.


It's been said there are only two stories: a man goes on a journey; a stranger comes to town. Of course, since any journey in which a man leaves one place will involve him ending up in another, and vice versa, we might say there's really only one story: the loss and discovery of the world.

The simple act of placing a person in an environment with which he's unfamiliar and surrounded by people whose ways are strange to him creates the most transparent means of describing that place and its customs, particularly in a story that doesn't rely on some metafictional conceit of narration or direct address to the reader. An eleven-year-old boy, upon waking up in the morning and coming downstairs for breakfast, doesn't think, "This is my mother, who was educated at Bryn Mawr and Penn and manages a hedge fund; this is my father, who met my mother at Penn and lost his job as a trader on Wall Street and now wants to become a mixed martial artist." He knows this already, so there's no reason for him to do so. But if the eleven-year-old has had a friend sleep over, especially a new friend, then his friend will ask about his parents, or find a trove of MMA videos and Bryn Mawr memorabilia, or, failing subtler devices, the protagonist will simply offer up the relevant information.

So: one story, acknowledging that what constitutes "journey," "stranger," and "town" are broad concepts, which is fortunate for contemporary writers, because it allows for innumerable iterations of this single theme. Additionally, history (which didn't end during the Reagan presidency after all) continues to supply us with new milieux, and indeed new places, to be explored and written about. Thus, moving beyond the aspects of setting, we encounter the types of setting. Perhaps a better term is "mode," since the choice of where and when to set a story impacts not just the physical world and milieu the writer has to create or invoke, but how he or she does so. Indeed, many genre distinctions are based less on the events or characters in a story than on the world in which it is set.

Of course there is the contemporary, local setting I mentioned at the beginning of this essay: the world we inhabit right now. And, though I made fun of it, the truth is that, in art as in most things, novelty is very often less important than how well the writer realizes his or her subject. Nowhere is this more evident than in memoirs or the contemporary bildungsroman, in which one story of a typical middle-class suburban childhood can be the most boring, irrelevant, and self-indulgent piece of twaddle you've ever read, whereas another equally typical, equally middle class, equally suburban childhood can be a revelation. There is the historical: Beloved. Blood Meridian. Waiting for the Barbarians. The novel of the future: 1984. Infinite Jest. The Walking Tour by Kathryn Davis. The novel in which certain aspects of the setting actually exist in the real world but others do not. These include so-called alternate-history novels such as The Man in the High Castle by Philip K. Dick, or Roth's The Plot Against America, or the magical realist novels of Garcia Marquez or Angela Carter; in this category we would include most science fiction, and a lot of children's fiction as well, be it Harry Potter's slant perspective on an otherwise banal reality, or C.S. Lewis's Chronicles of Narnia, which posits an extension of our world only tangentially connected to it through hidden portals. Then there are the novels of purely fictitious universes bearing no physical or temporal connection to the one we live in—a task so daunting that it has, paradoxically, become the province of the worst rather than the best writers, as anyone who's ever perused the fantasy section of a bookstore will know. One of the great twentieth-century tropes is the world that is completely or partially the product of a character's imagination: a dream, as in The Wizard of Oz; or madness, as in David Markson's Wittgenstein's Mistress; or complete self-absorption, as in Knut Hamsun's Hunger or Beckett's great trilogy of 1946-50. One of the pleasures of teaching my class on setting is that I get to read and reread these and other books—books like Ovid's Metamorphoses (and Kafka's Metamorphosis), Wuthering Heights, Benito Cereno, A Book of Common Prayer, The Waves, The Rings of Saturn, Wide Sargasso Sea, all novels in which the setting—the world—far from being taken for granted, far from being a purely objective or purely subjective phenomenon, exists in a complex and singular relationship with its characters. At the beginning of this essay I referred to that relationship as narrative, but each time I finish one of these books I am reminded—reassured, I should say—that what I really mean is life.


I began teaching my class with the intention of demonstrating that narrative is nothing more than—indeed, can only be—a byproduct of setting and character. In fact, this relationship strikes me as so obvious that I often wonder why no one's suggested it before, or, if someone has and I've just missed it, why it's not more widely accepted.

Well. When in doubt, turn to the Greeks. In the Poetics, Aristotle devotes the bulk of his attention to narrative and character, with "spectacle" coming a distant fifth and a half (after reasoning, diction, and a passing nod to song). Aristotle was speaking of poetry in general and more specifically of drama, but his comments about the relationship of narrative and character, and how they relate to the world in which they take place, still have profound implications for today's lowly prose writers.

For Aristotle, the classic combo of pity and fear that creates tragedy comes from actions rather than qualities; hence narrative's supremacy. To state that a man is noble or wicked or misguided or a pawn of fate is important but uncompelling. One must show how he is noble, wicked, misguided, etc., and that requires action. Aristotle was speaking to a story's effectiveness, or, we might say, affectiveness: the ability to produce the all-important catharsis, and perhaps impart some measure of education or edification as well.

The essence of Aristotelian narrative is plot, an ordered sequence of events that produces an intended effect, emotional (cathartic) and intellectual (pedagogical). To clarify: narrative is any sequence of events in any order about any number of people; plot is an ordered sequence of events in which chronology, character, and the events themselves have been carefully presented so as to produce an emotional and intellectual response in its audience. In Greek drama as well as in twenty-first century fiction, the ordering intelligence is the author's, but in Greek drama as opposed to twenty-first century fiction, this order mirrored a religious belief that the events of each of our lives have been arranged by a supernatural intelligence. Mimesis, or imitation, reproduced not just the Greeks' physical but also the spiritual world—or, more specifically, the spiritual world's control over the physical world—whereas contemporary writers must satisfy themselves with the material universe and its visible or quantifiable phenomena.

This worldview stands in sharp contrast to the secular worldview mentioned earlier, but it's important to remember that the teleology is Greek, not Judeo-Christian (sorry, Herr Liebniz). The Olympian gods controlled their believers' destinies and meddled in their lives not for the sake of some edifying moral purpose, but simply because they could. Shakespeare said we were all players on a stage; in the Greek world, the hooting, hollering, and otherwise amused, bored, and profoundly fickle audience was clearly the gods. Seen in this light, the Greek admonition to "know thyself," rather than a paean to self-improvement, becomes instead a mere instruction to learn who one already is, and accept one's place in the grand scheme of things. If you are a slave, serve your master faithfully. If you are a woman, serve your husband. If you are poor, work.  If you are rich, build monuments. If you are noble, rule. Aristotle gives modern readers something of an out in that he considers intention to be a kind of action. A slave can, indeed, yearn to be free—but it usually turns out that he yearns for freedom because he is in fact a king or prince wrongfully enslaved.

So: caste equals character. Anatomy is destiny. The wicked can never be reformed and the just can never commit a crime, save under false pretenses or duress of provocation. People can never change who they are, only how they are. It could be said that the entirety of western civilization has been a struggle to overcome this kind of fatalism—or, to use the contemporary term, essentialism. It even has a preferred narrative, namely, that of the self rebelling against the strictures of a society that prohibits its free or full expression—which is reminiscent of what was observed earlier about milieu, namely, that just as writers must engage with but also subvert readerly expectations of who people are and how they behave, so too must their characters struggle to overcome the social pressure to conform.

Aristotle believed that among certain highly-prized innate qualities—athleticism, a good singing voice, nobility—was the ability to make profound, illustrative comparisons—i.e., metaphor, which in turn was the cornerstone of all great literature. Like a runner or a singer, you can hone your God-given talents, but ultimately you're either born an artist or you're not. I'd like to look at a single piece of writing a little more systematically in order to show how this idea of metaphor has been bound up with the ways writers create their fictional worlds since literature's very beginnings, because it moves us from craft to exegesis, from how writers create and manipulate setting to the more important question raised at the beginning of this essay, namely, what is the relationship between the world readers encounter in fiction, and the one they live in?

I am of course talking about The Iliad. Homer's text is rife with hints and half-truths and facts as devoid of context as an unsigned letter found in a used book. As an historical document, it was accurate enough to guide Heinrich Schliemann to Troy thousands of years after the city was destroyed, but it also features gods and goddesses—and nymphs, centaurs, and river spirits—fighting side by side with human soldiers, all of which calls into question the degree to which Homer conceived of himself as documenting an actual episode in the early history of the Hellenes, and the degree to which he was filling in gaps in the historical record with imaginative mythmaking. One of the few things we can say, however, is that the conventions of storytelling were more or less fully formed by the time Homer put together his epic. A clean syntactical line, a modulated but generally fast-moving plot, a cast of diverse characters at the heart of which is a brooding antihero whose motivations are as complex as any of contemporary fiction's cardboard Christs: Homer utilizes these conventions proficiently, in the manner of a craftsman, rather than ostentatiously, in the manner of an inventor. They unify his 15,000-line poem and keep it, among other things, intensely, pleasurably readable for a modern audience. In fact, the same tools Homer used 2,700 years ago continue to shape today's novels, which, on balance, have more similarities to The Iliad than differences. They remind us first of all that storytelling is as old as civilization, if not simply a precondition for it; but also that Homer, though he may or may not have been the first poet to write a story down, clearly wasn't the first to tell one.

What does feel new is metaphor, at least in this iteration. Indeed, we seem to be witnessing its birth, a manifestation as explosive as Athena's emergence from her father's skull. Zeus, the stories tell us, suffered from an acute headache before Athena finally split his brow and strode forth; and metaphor seems for much of The Iliad similarly painful to both Homer and his audience. I mean here the ornamental epithets, the "wine-dark seas" and "long-haired Achaeans," which serve less as description than metrical and mnemonic convention. Though some of them are beautiful ("rosy-fingered dawn" remains a favorite), many are nearly invisible ("white-armed Hera," "swift-footed Achilles," "sea-going ships"), and do little more than fog the language as they fill a metrical gap; few if any pass the Aristotelian test of profundity.

Then there are the belabored and often imprecise similes that liken virtually all aspects of soldiery—stealth and pursuit, clash and flight—to the habits of the lion, or the bull, or the snake:

But over against him came Achilles rearing like some lionout on a rampage, and a whole town of men has gearedfor the hunt to cut him down: but at first he lopes along,all contempt, till one of the fast young hunters spears him—then...crouched for attack, his jaws gaping, over his teeththe foam breaks out, deep in his chest the brave heart groans,he lashes his ribs, his flanks, and hips with his tail,he whips himself into fighting-fury, eyes glaring,hurls himself head on—kill one of the men or die,go down himself at the first lethal charge!So now magnificent pride and fury lashed Achilles....
Well, yes. But also no. Achilles isn't wounded, nor does he have any fear of death. Nor, for that matter, is he "out on a rampage": he's engaged on a specific mission. What most strikes one about these passages, especially as they accumulate over the course of the text, is their seeming desire to become stories of their own—and Homer's abandonment of them at just that point. It's as if Homer is rejecting the idea that metaphor is another kind of narrative, as if he's pushing his comparisons to see if they can yield something other than story.

This search comes to sudden—one wants to say unplanned—fruition in Book 18, when in an abrupt shift of scene the poet abandons his battlefield narrative for the heavenly wrought shield of Achilles. "With all his craft and cunning the god creates a world of gorgeous immortal work," Homer writes. Hephaestus's shield is so godlike that it doesn't merely depict the world: it enacts it. When farmers till a field "the earth churned black behind them, like earth churning, solid gold as it was—that was the wonder of Hephaestus's work." What makes this labor even more astonishing is its irrelevance. Achilles's mother Thetis has asked Hephaestus to make her son a new suit of armor so he can face Hector, who killed Achilles's beloved Patroclus. But Thetis's request is split in two by a curious aside: "I throw myself at your knees, please help me! Give my son—he won't live long—a shield and helmet and tooled greaves with ankle straps and armor for his chest."

He won't live long: in fact, Achilles's death is fated to follow "hard on the heels of Hector's," and anything that aids that victory abets Achilles's ultimate defeat. Hephaestus acknowledges this when he replies, "Anguish for all that armor—sweep it from your mind. If only I could hide him away from pain and death, that day his grim destiny comes to take Achilles, as surely as glorious armor shall be his...." Nevertheless the god of the forge fires up his twenty bellows, gathers his bronze and gold and silver, his hammer and tongs, and labors on the "indestructible" yet useless shield for 141 lines—lines devoted not so much to a description of the shield as to the making of it. Compare this rhetorical glut to the breastplate that does nothing more than shine "brighter than gleaming fire," or Thetis, who can only flash "like a hawk" as she hurries these gifts to her doomed son. Clearly, the difference is not simply one of scale.

No, for the first time in the poem, comparison opens into something larger. Call it correspondence, call it symbol, but Hephaestus's shield isn't simply like the world: it is the world. There is no "like," there is no "thing that lies beneath the semblance of the thing" that Virginia Woolf searched for in The Waves. Hephaestus's shield is not a metaphor for the world, but a perfect act of mimesis. The god of fire and forge does what Homer and all human artists aspire to do but cannot. He makes a copy of the world so perfect that it becomes real—so perfect that it eclipses what it emulates and leads to the death of the person it ostensibly protects. But note the process by which Homer reveals this terrible truth about art: a metaphor that pretends to be a story is misleading and should be abandoned—all those bears and snakes and lions. It is story that is the source of metaphor, not the other way around. The real world, in other words, cannot be compared to an artist's creation, whereas an artist's creation stands as nothing more than a comparison to the real world—one that always comes up wanting.

This is a paradigm that artists have been struggling with ever since. Hence postmodernism's frequent jabs at the fictional dream; hence the general decline in interest in the novel as opposed to nonfiction. If the inventions of men can never equal the perfection or permanence of the work of God, then why bother? But is that really what the bard is saying? Look at Hephaestus's shield in action. Its power hardly comes across as absolute. A single thrust of Aeneas's spear "bore through two plies" of the five that make up the shield; when the river Scamander turns on Achilles, Hephaestus's gift is actually a hindrance as the water "slammed against his shield and he staggered, lost his footing"; when the Xanthus River joins the Scamander in the attack, he boasts that "all that glorious armor" will sink "somewhere under our floods," and it's clear the rivers would have made good on their threat had not Hephaestus used his fire against them—the same fire that made Achilles's shield. Hephaestus's creation, then, reflects the world so perfectly that it paradoxically makes the shield as imperfect and impermanent as the world it emulates. What is inexhaustible is the power that made it. What Homer is discovering for future storytellers is that metaphor's creations don't simply imitate life. They also efface it, and are in turn effaced by it. Metaphor, the heart of every story, exposes the lie of storytelling's supposed goal of protecting the past. As language's progenitive monument, metaphor, like the sex drive itself, safeguards only the future. Hephaestus's shield doesn't save Achilles because it's not meant to. It seals his doom, but in so doing protects something more than a hero or, for that matter, the memory of one. Achilles has died, and his shield is long since lost to history. Only the process that made it lives on. Indeed, Homer's forge has proved more lasting than the gods who fired it.

Earlier I wrote that I began these observations in response to a perceived limitation in my students' writing. But there was another, more pointed context as well: September 11. As the horrific yet beautiful image of the smoking towers glowed in the nation's eye and the noxious odor of crematorium fires wafted through the city—my city—I wondered, like many writers, if I should write about such a tragedy, and if so, what should be written. I turned to literature for possible models. I read "Lycidas" ("weep no more, for Lycidas your sorrow is not dead"), I read the philosopher's monologue at the end of The Metamorphoses ("all things change, yet never die") and then, finally, I read Homer's catalogues of the dead, the Greek and Trojan soldiers who fell by the thousands at the point of each other's spears, and whose deaths Homer enumerates as faithfully as the Torah catalogues the births of the first families. But as I worked my way through these monumental literary responses to death, I came to realize I was looking less for a model than for solace, a context for my grief. I read not as a writer, but as a reader overwhelmed by the open-ended shapelessness of a tragedy. Perhaps it takes something like 9/11 to remind us that literature has qualities beyond the aesthetic; perhaps that's merely the kind of sentimentality that surfaces in times of duress. But as I read how "Ajax brought down Hyrtius, Gyrtius's son, a lord of the ironhearted Mysians," how "Antilochus slaughtered Phalces" and "Meriones killed off Morys," how "Teucer cut down Periphetes and Prothoon," and "Menelaus took the hardened captain Hyperenor, gouged his flank and the bronze ripped him open, spurting his entrails out," I recognized that, whatever I wrote, it didn't need to be a monument, because these lists—Homer's lists—already adorned scaffolding and light poles throughout the city. National sentiment had transformed our victims to heroes well in advance of any literary myth-making, and from these heroes' ashes America's new war had sprung fully formed, as implacable as Athena and insatiable as the phoenix. This New York—this America—required no aesthetic intervention, from me any more than my wide-eyed students, in order to be felt. It needed to be experienced rather than reproduced, for its singularity to be affirmed rather than capitalized upon. Not "What happened?" but "What happens next?"

Our moral duty as writers is not to mourn or make monuments to worlds that, by dint of the passing of time as much as any more nefarious acts, are always being lost. As Leon Wieseltier admonished Martin Amis when he reviewed Amis's collection of essays on 9/11 and its aftermath: "When he describes the second plane on its way to the south tower as 'sharking in low over the Statue of Liberty,' the ingenuity of the image is an interruption of attention, an ostentatious metaphorical digression from the enormity that it is preparing to reveal, an invitation to behold the prose and not the plane."

The plane is the real world, the prose our imitation of it. Medieval Christians dramatized a similar relationship between the earthly and spiritual realms by appropriating the ancient Pythagorean symbol of the vesica piscis, also known as the mandorla:

The mandorla is the almond-shaped central section formed by the overlap between the two circles—the two shields I want to say, Hephaestus's and Homer's, arrayed in phalanx formation—and its proportions are precisely fixed; the center of each circle lies on the circumference of the other, meaning that there is no more holy or less holy world, no special circumstances under which we grow closer to or farther from heaven's perfection. The mandorla is a measure both of God's presence in our life, and how far removed he is as well; of how wondrously made are certain aspects of creation, how chaotic and incomprehensible others are; the best of all possible worlds, and an acknowledgement of just how conditional and unsatisfactory that statement is. Revered figures were depicted within the protective aureola of the mandorla to symbolize their dual nature, and, as a writer, I've come to think of it as a symbol for the relationship between the worlds we make in our fiction and the world we model our creations on—indeed, the two sides of the mandorla look like parentheses, giving what they contain an appropriately contingent quality.

When you read, remember that everything you encounter is by its very nature derivative, an imitation rather than a thing in itself, no matter how carefully each scene is crafted, each sentence is wrought. Heaven; earth. Setting; character. The plane; the prose. Never forget where you should be looking.

All artwork by Yohei Oishi.