A Mortal Voice for an Immortal Monster

The Life of a Goddess that is not a Heroe's Memory

Every epic keeps a graveyard of minor women just offstage: the nymph who delays the hero for a chapter, the witch who tests him, the queen who waits at the loom while history happens elsewhere. Madeline Miller's second novel is an act of exhumation. It takes one such figure, the enchantress of Aiaia who turns Odysseus's crew to pigs in the tenth book of the Odyssey and is never heard from again once the ship sails on, and gives her a birth, a girlhood, a crime she cannot undo, a species of motherhood no one warned her about, and an old age she has to choose for herself, since immortality does not come with one built in. The wager of the book is that a life lived entirely in someone else's footnotes might, examined closely enough, turn out to contain a whole epic of its own. It is not a modest wager, and Miller, to her considerable credit, mostly wins it.

What is Circe attempting to do? Something more ambitious than it first appears, dressed as it is in the accessible clothes of a page-turning, single-narrator retelling. It wants to take a figure whom Homer, Ovid, and two and a half thousand years of retelling have used chiefly as an instrument, a test administered to men, a danger for heroes to survive, and grant her the one thing instruments are never given: an interior. Does it succeed? For some three hundred pages, remarkably so, in a voice that manages the difficult trick of sounding ancient and legible at once. In its final stretch, the novel's nerve falters somewhat, trading the harder discipline of consequence for the softer comforts of a fantasy made good. Why does it matter? Because Circe arrived at the crest of a real and significant literary movement, the systematic re-narration of the classical world from its margins, and did more than any single book since to prove that movement commercially formidable as well as critically serious. A witch who was, for two and a half millennia, permitted only to speak in the third person has, at last, been given a mortal's voice. What she says with it is the substance of what follows.

The bones of the plot will be half-familiar to anyone who has read Book Ten, and wholly unfamiliar to everyone else. Circe, least-loved daughter of the sun god Helios and the scheming naiad Perse, is a disappointment from birth: not beautiful in the golden family style, not gifted with her mother's cunning, possessed of an unaccountably human voice that marks her as strange among gods who sound, to mortal ears, like breaking rock. She falls first for a mortal fisherman, Glaucos, and transforms him into a god using flowers grown from the blood of the castrated Titan Kronos, then, in a jealous rage when he abandons her for the nymph Scylla, transforms her rival into the six-headed monster of the Odyssey's most famous strait. For this she is exiled to the island of Aiaia, where she teaches herself witchcraft from nothing, is raped by a crew of sailors and thereafter turns every subsequent crew of men who threaten her into pigs, hosts Daedalus and midwifes the Minotaur's birth for her sister Pasiphaë, entertains Hermes and eventually Odysseus for a year, bears a son by him whom she raises alone against the enmity of Athena, and at the novel's close chooses, after millennia of divinity, to become mortal. It is, in outline, the whole of Greek myth's minor circuitry rewired to run through a single consciousness. That it mostly holds together is the first thing worth admiring about it.

The Voice That Learns to Speak

The novel's governing formal decision, and its single greatest technical achievement, is the sustained first-person voice across a timescale no mortal narrator could survive. Miller opens on the day of Circe's birth and does not relinquish her point of view for the whole of what must be several thousand years, a structural gamble that could easily have produced either numbing sameness or unreadable compression. Instead, she solves it with an unglamorous but effective trick: narrative pacing keyed not to elapsed time but to felt significance. Centuries of Circe's childhood at her father's feet pass in a single elegant paragraph, tallied off almost as a list, while a single afternoon with Glaucos on a beach can run several pages, because the girl experiencing it has no way yet of knowing which moments will matter. It is, in its quiet way, a formally intelligent solution to a problem most immortal-narrator novels do not even think to pose themselves.

The prose favors clean, declarative sentences with the occasional Homeric flourish held in reserve for moments that earn it, an epithet dropped here, a formal repetition there, never overused. Miller's ear, trained by years spent teaching the ancient languages rather than merely reading translations of them, shows in small structural choices a less disciplined writer would miss: the way divine speech in the novel tends toward pronouncement and command, all subject and verb, while Circe's own interior voice, particularly once she has spent real time among mortals, grows steadily more qualified, more hedged, more willing to sit inside a doubt rather than resolve it immediately. The prose is, in other words, doing quiet characterization work at the level of syntax, charting Circe's slow humanization not through what she says she feels but through how her sentences are built.

Where the novel is less disciplined is in its handling of secondary voices. Because everything is filtered through Circe's retrospective telling, characters she does not fully understand tend to remain slightly flat on the page regardless of how much the plot needs them, Daedalus most notably, whose grief over Icarus is reported to devastating effect but whose interior life before that grief never quite escapes the function Circe assigns him, which is craftsman, confidant, and eventually lover. Odysseus fares somewhat better, partly because Miller gives him enough contradictory self-report, the boastful war stories, the private admissions of fear, the chilling justification of a child's murder as necessary statecraft, that the reader can triangulate a fuller man than Circe herself may fully see. It is a case where the novel's formal constraint, a single unreliable rememberer, becomes a narrative asset rather than a limitation, at least in the hands that know how to use it.

From Ovid's Footnote to a Generation's Heroine

Circe enters the Western tradition already secondhand. Homer gives her some four hundred lines in Book Ten of the Odyssey, a dangerous hostess who drugs Odysseus's men into swine before the hero, protected by the herb moly that Hermes presses into his hand, forces her to relent and becomes her lover and guest for a year. Hesiod's Theogony supplies her genealogy, daughter of Helios and the Oceanid Perse, sister to Aeetes and Pasiphaë. Ovid, in the fourteenth book of the Metamorphoses, supplies the darker material Miller draws on most heavily and reworks most substantially: the jilted-lover transformation of Scylla into a monster, born there of spurned desire rather than, as Miller frames it, something closer to displaced grief and rage at her own repeated powerlessness. A lost epic called the Telegony, surviving only in fragments and summary, supplies the material of the novel's final act, Circe's son by Odysseus, and the son's unwitting killing of his own father with a spear tipped in the venom of a stingray, a detail so strange that Miller's decision to include it rather than soften it is itself a small act of critical courage.

What Miller does with this scattered inheritance is not, strictly speaking, invention so much as synthesis and interiorization, gathering material from four or five source texts spanning nearly a thousand years of antiquity and welding it into a single continuous consciousness that none of those texts, taken separately, ever offered. This is itself an ancient and honorable technique, closer to what a fifth-century Athenian tragedian did with inherited myth than to what a contemporary novelist typically does with a blank page. Aeschylus did not invent Prometheus; he built a play around the god already chained to his rock. Miller does not invent Circe; she builds a consciousness around the witch already waiting on Aiaia. The novel's opening set piece, Circe as a girl witnessing Prometheus's punishment and bringing him water in secret, is in fact the clearest signal the book offers of its own methodology: an act of small, unauthorized compassion extended toward a figure the official story has already condemned and moved past.

There is also a nearer and less frequently noted ancestor: the mid-century historical novel of classical antiquity practiced by Mary Renault, whose The King Must Die and its sequels perform a comparable act of psychological ventriloquism for figures like Theseus, granting interior plausibility to men the myths had rendered as pure action. Miller's innovation, against that tradition, is to point the same lens at a woman the myths rendered as pure danger, and the shift in target is not incidental. Renault's project rescued heroes from cardboard. Miller's project rescues a monster from being merely a monster, which is a harder and, in the current moment, a considerably more urgent kind of rescue to attempt.

Immanence, Transcendence, and the Trouble with Truest Selves

Simone de Beauvoir's account of woman as historically confined to immanence, to the repetitive, unauthored maintenance of life rather than the world-making, self-defining projects reserved for men, reads as though it might have been written with the opening chapters of this novel open on the desk. The nymphs of Circe's girlhood exist in a condition of pure immanence: they weave, they bathe, they arrange themselves to be looked at, they wait to be chosen in marriage by whichever god or hero finds them useful, and the text is admirably precise about how suffocating this is meant to feel, not through complaint but through the sheer numbing repetition of Circe's early chapters, centuries reduced to a paragraph because nothing in them required a self to happen. Witchcraft, when Circe finally discovers it, functions in exactly Beauvoir's terms as an act of transcendence: for the first time in her existence, Circe undertakes a project of her own authorship, one whose outcome is not given in advance by her divine nature but must be labored for, failed at, and won through effort that resembles nothing so much as a craft apprenticeship. The acorn she rubs with oils for months and gets nothing from is the clearest image in the book of transcendence as something earned rather than granted.

Yet the novel's central magical premise sits in real and, I think, largely unexamined tension with this existentialist frame. The flowers that transform Glaucos and, later, Scylla are said to reveal a creature's truest self, an essentialist claim, the idea that each being possesses one fixed, authentic nature merely waiting to be released, which is philosophically almost the precise opposite of the self-authorship Circe's own witchcraft otherwise celebrates. If Circe is a existentialist heroine, a self made through chosen action rather than discovered through fixed nature, then her signature act of power operates on a metaphysics she herself is in the process of disproving. The novel gestures toward this contradiction exactly once, when Aeetes suggests that what the flowers reveal as Scylla's truest self looks suspiciously like what Circe's own jealousy wanted to find there, and then largely drops the thread rather than following it to its uncomfortable conclusion.

The novel's closing philosophy, meanwhile, draws closer to Heidegger than to Beauvoir, though I doubt Miller would object to the company. Circe's culminating insight, that the gods, immune to death, are in some sense more lifeless than mortals precisely because nothing is ever truly at stake for them, nothing chosen against the pressure of an ending, is a fair novelistic rendering of the Heideggerian claim that authentic existence requires being-toward-death, a horizon against which choices acquire weight. A being who cannot die cannot, in this account, truly choose, since every choice can simply be revisited across eternity without cost. Circe's final decision to drink the mortality draught is therefore not a surrender but the first fully existentialist act available to a god: the deliberate acceptance of a horizon, and with it, finally, the possibility of a life that means something because it will not last forever. It is the novel's most successful marriage of ancient material to modern philosophical concern, and very nearly redeems the metaphysical carelessness of the flowers that made Scylla a monster in the first place.

What the Witch Never Has to Answer For

The first and most consequential blind spot concerns the moral accounting of the pig transformations themselves. The origin of the practice is unambiguous and devastating: Circe is raped by the captain of a visiting crew, and her subsequent, near-automatic transformation of every threatening man who lands on her shore thereafter is presented, correctly, as a hard-won act of self-defense rather than casual cruelty. But the novel sustains this practice across what appears to be centuries, transforming crew after crew, and the text's moral temperature toward these later transformations remains oddly untroubled. We are told Circe occasionally lets pious, harmless men go, which acknowledges some discrimination is happening, but we are never made to sit with the fact that she has become, structurally, precisely the kind of casually violent power she spent her exile despising in her own family, dispensing punishment by instinct rather than judgment, exactly as her father and uncles once did to her. A novel this alert to the abuse of power elsewhere declines to ask its heroine the one question it asks everyone else.

The second concerns Scylla, and here the blind spot is one of narrative economy rather than moral evasion. Circe creates a monster who will go on, by her own late estimate, to kill sailors indefinitely, a horror on a scale that dwarfs anything the gods around her are shown doing, and the novel treats this largely as personal tragedy, a wound Circe carries, rather than as an ongoing atrocity for which she bears direct, continuing responsibility to people she will never meet. The line that transformation touches only bodies and not minds is one of the book's sharpest observations, but it is deployed almost entirely in service of Circe's own guilt rather than in service of imagining the actual, ongoing cost to the sailors who will die in that strait for millennia after the novel's final page.

The third is the marked softening of the novel's final act. Having spent three hundred pages establishing a world governed by patriarchal cruelty, casual sexual violence, and the structural powerlessness of nymphs against gods and men alike, the book resolves into a fairly conventional domestic fantasy: a good marriage to Telemachus, imagined daughters, a vision of quiet family happiness rendered in a dreamlike present tense whose reality within the fiction is left deliberately ambiguous. This ambiguity may be intentional, a final wish rather than a final fact, but the effect, whether by design or not, is to let the novel's harder, angrier material recede behind a closing image that a more conventional romance novel might have supplied with equal ease.

The fourth is the flattening of scale that comes with compressing millennia into a single continuous voice. The novel is, in its way, guilty of the very indifference to duration that it criticizes in the gods, folding centuries into paragraphs when convenient and stretching single afternoons into chapters when they serve the emotional throughline, a narrative choice that is defensible as craft but that goes curiously unexamined as theme, given how directly the book's own philosophy turns on the difference between mortal and immortal relationships to time.

What Miller Accomplishes

It would be an ungenerous critic who let the ledger above obscure the scale of what this novel accomplishes with difficult raw material. The sustained first-person voice across millennia, discussed earlier as formal risk, succeeds far more often than it fails, and succeeding at all is itself an achievement most writers would not attempt. The handling of Scylla, whatever its later narrative neglect, is in its immediate telling a piece of morally serious storytelling rare in popular fiction: a heroine who does something permanently wrong, whose magic cannot undo the harm, and who is made to carry that irreversibility rather than have it conveniently reversed by a later act of grace. Too many novels built around a sympathetic protagonist would have found a way to fix Scylla by the final act. Miller, to her lasting credit, does not.

The refusal to soften Odysseus is likewise an act of literary nerve. A lesser novel, especially one narrated by a woman who loves him, would have sanded down the war stories, the murdered Palamedes, the abandoned Philoctetes, the slaughtered sleeping men, the child Astyanax thrown from the walls and justified as necessary statecraft. Miller lets Circe love him anyway, clear-eyed about what that love costs her judgment, and the tension between admiration and horror that results is more honest about how love actually survives knowledge of a person's cruelty than almost anything else the book attempts.

Structurally, the decision to build the entire novel from scattered, sometimes contradictory ancient sources, Homer, Hesiod, Ovid, the fragmentary Telegony, and weld them into continuous psychological cause and effect, deserves recognition as a feat of synthetic craft, closer to the work of a skilled dramaturge assembling a coherent production from conflicting stage directions than to ordinary invention. The moly plant that saves Odysseus in Homer becomes, in Miller's hands, a quiet running motif for the limits of Circe's own power, present at both the novel's midpoint and its climactic confrontation with Apollo, threaded through with a discipline that rewards attentive rereading.

On What Circe Does to Scylla, and What Miller Does to Circe

Set beside its nearest contemporaries, the novel's particular achievement comes into sharper focus. Pat Barker's Briseis, in The Silence of the Girls, is angrier and less consoled, a voice that refuses the reader the comfort Miller's Circe eventually extends. Christa Wolf's Cassandra is stranger and more formally fractured, using myth to say something pointed about a specific totalitarian present that Miller's book, more interested in the timeless structures of patriarchy than in any single political moment, does not attempt. Judged against Miller's own earlier novel, The Song of Achilles, Circe is the more ambitious achievement by any fair measure, trading a single tight tragic arc for the far harder problem of sustaining a life across the whole of recorded myth, and mostly solving it. Judged against the classical sources it draws from, the novel is generous almost to a fault, softening Ovid's coolly clinical cruelty into something warmer and more explicable than the ancient material ever offered or perhaps even intended.

But the most interesting thing this novel does may be a thing it never quite says aloud, and it lies in the very magic it gives its heroine. Circe's flowers, we are told, reveal a creature's truest self: they take a fisherman and show the god hidden inside him, they take a jealous rival and show, or so Circe tells herself, the monster hidden inside her. This is not merely magic. It is, with only the thinnest disguise, the operation of a novelist upon her source material. Miller takes a minor figure from someone else's story, a witch granted four hundred lines and a plot function, and decides what was truly inside her all along: a girlhood, a wound, a conscience, a capacity for growth Homer never thought to grant her. She hands that truest self back to the reader, transformed, as a gift the ancient Circe was never asked whether she wanted.

Whether that transaction, performed here with real skill and evident love for the source material, deserves to be called liberation, or something closer to a beautifully executed act of authorial appropriation, no less an imposition of a truest self than the one Circe forces on Scylla, is a harder question than the one the novel spends four hundred pages answering, which is simply whether its heroine deserved a voice. She did, unquestionably. But the reader finishes wondering whether the voice she has been given is entirely her own, or whether it belongs, in the end, to the twenty-first century that wanted a witch it could love, and reached back two and a half thousand years to make one.