Adaptation Is Not Fidelity: How to Make Old Texts Feel Dangerous Again

The strongest book to film adaptation does not preserve a novel like evidence under glass. It identifies the force that once made the story volatile, then rebuilds that force through screenplay adaptation, cinematic form and, when necessary, radical reinvention.

A strange thing happens to a famous book when it becomes respectable. The sentences survive. The plot survives. The characters become cultural furniture. What often dies is the threat.

Section Index:

• The Problem With Fidelity
• Event, Function, Effect
• Preserve the Force, Not the Costume
• When Expansion Is the Right Move
• Re-Authoring Point of View
• A Hierarchy of Fidelity
• Audience Memory as Screenplay Machinery
• Compression Without Losing Complexity
• A Practical Workflow for Adapting Literature
• Danger Versus Reassurance

The Problem With Fidelity

The question is newly visible because adaptation culture keeps returning to the same canonical wells. In September 2026, Georgia Oakley’s new Sense and Sensibility, written by Diana Reid, arrived with an explicitly recalibrated emphasis on the Dashwood sisters as mutually formative rather than as simple embodiments of opposing temperaments.^[21] Whether that interpretation ultimately convinces is less important here than the premise behind it: another Austen adaptation cannot justify itself merely by being newer. It has to decide what the novel’s old conflict means under a new gaze.

That is the central problem of adapting literature for the screen. It is not primarily a matter of shortening 400 pages into 110, nor of deciding which beloved scene can be sacrificed without enraging readers. Those are logistical problems. The artistic problem is more severe: how do you make an audience feel uncertainty, danger, erotic charge, social pressure or moral disorientation when the story has already been read, taught, quoted, merchandised, filmed and remembered?

The conventional answer is fidelity. Preserve the story. Preserve the dialogue. Preserve the period. Preserve the author’s intention, insofar as anyone can confidently reconstruct it. Yet adaptation theory has spent decades exposing the weakness of that standard. Linda Hutcheon’s useful formula is that adaptation is “repetition without replication”: a work returns, but in another medium, another context, another historical moment and through another artist’s interpretation.^[1] Robert Stam went further, arguing that the language surrounding supposedly unfaithful adaptations has often been suspiciously moralistic, full of betrayal, violation and desecration, as if the novel were a sacred object and the film a vandal.^[2]

But rejecting fidelity as the supreme value creates another problem. If an adapter owes nothing to the source, why adapt it at all? Why not write something else?

The answer is that a serious book to film adaptation owes the source something more demanding than obedience. It owes it an interpretation.

The screenwriter’s task is to identify the source text’s pressure system: the arrangement of desires, taboos, social forces, mysteries, humiliations and contradictions that makes the work emotionally unstable. Then the screenplay must reproduce that instability by cinematic means. Plot may survive. Plot may also be the first thing that needs to be broken.

Event, Function, Effect

Thomas Leitch’s survey of adaptation studies describes a field repeatedly reorganized around disputes over fidelity, medium specificity and intertextuality.^[3] André Bazin had already treated literature and cinema less as master and servant than as distinct aesthetic systems whose collision could reveal what each medium actually does.^[4] The most useful practical consequence is simple: the adapter should stop asking, “How do I put this book on screen?” and start asking, “What experience did this book produce, and what would cinema have to do to produce an equivalent experience now?”

That distinction changes everything.

A novel can imprison us in consciousness for twenty pages. A film has a face, a room, a cut, an offscreen sound, a silence. A novel can stop time for commentary. A film can make three seconds unbearable. A novel can tell us that a character has misunderstood herself. A film may need to stage the misunderstanding as behavior, architecture, performance or contradiction. The source offers material. The medium demands decisions.

Recent scholarship on David Cronenberg’s Naked Lunch makes the point unusually starkly. A 2026 article in Adaptation argues that the supposed “unfilmability” of William Burroughs’s novel is not a mystical quality but a problem bound up with the different technical conditions of typewritten language and filmed image.^[5] Cronenberg did not solve that problem by illustrating Burroughs. He invented a cinematic object that mixes biography, hallucination, writing and grotesque machinery. The question was not how to film every incident. It was what cinema could substitute for the book’s destabilizing textual machinery. This is where adaptation stops being transcription and becomes dramaturgy.

A screenplay adaptation should begin by separating the source into at least three layers: event, function and effect. Event is what happens. Function is what the event does in the story. Effect is what it does to the audience.

Consider the difference. A mother burns a manuscript. That is event. The destruction may rupture a relationship and force the protagonist toward independence. That is function. The audience may feel that domestic love has suddenly revealed an authoritarian underside. That is effect.

Weak adaptation preserves the event because readers remember it. Competent adaptation preserves the function. Great adaptation asks whether the original effect can still be achieved with either. That last question is where old material becomes dangerous again.

Preserve the Force, Not the Costume

Akira Kurosawa’s Throne of Blood does not derive its power from presenting Macbeth with Japanese costumes. It relocates Shakespeare’s tragedy into feudal Japan and fuses it with formal elements of Noh theatre. The BFI describes the film’s Noh influence as present before the characters even appear, in its music, rhythm and atmosphere.^[6] The transfer is therefore deeper than setting. Shakespearean verbal density is not treated as the essence to be protected. Kurosawa finds another system for fatalism: fog, space, stillness, ritualized movement, faces that harden into masks, an environment that seems to have already decided the characters’ fate. This is a useful rule for modernizing classic stories: preserve the governing force, not necessarily its original costume.

The governing force in Emma, for example, is not Regency furniture. It is a privileged young woman’s confidence that she can read and arrange the social world better than the people living inside it. Amy Heckerling recognized that the mechanism could operate perfectly well in 1990s Beverly Hills. She has said that Austen’s novel felt “not at all dated” in its underlying social logic, even as Clueless rebuilt the world through high-school hierarchy, mall culture, slang, fashion and adolescent matchmaking.^[7] The adaptation works because Cher does not imitate Emma Woodhouse superficially. She occupies an equivalent social machine.

That phrase, equivalent social machine, is more useful to a screenwriter than “faithful update.” It asks concrete questions. What now performs the role that inheritance once performed? What institution now produces the same shame? What contemporary behavior carries the danger once attached to a letter, a duel, a ruined reputation, a forbidden marriage, a religious transgression? Which social fact has changed, and which human appetite has not?

Modernization fails when it updates nouns but not consequences. Give the heroine a smartphone, replace the carriage with an SUV, let everyone swear, and the result may still feel embalmed because the underlying power relations have not been reimagined.

Park Chan-wook’s The Handmaiden demonstrates the opposite approach. Sarah Waters’s Fingersmith is moved from Victorian Britain to Korea under Japanese colonial rule. The film retains a three-part structure and the machinery of fraud, confinement, class and erotic discovery, but its new historical setting changes the charge running through those elements. BFI criticism emphasizes how Japanese occupation, architecture, language, sexual power and class become active components of the film rather than decorative substitutions.^[8]

The change of setting is therefore an argument. It says that the source’s machinery of domination can be made newly legible by passing it through a different history. That is what interpretation looks like on screen.

When Expansion Is the Right Move

It also explains why screenplay adaptation should not automatically worship plot efficiency. Much practical screenwriting advice is correct about compression: films often require subplots to disappear, characters to merge and interior thought to become behavior. A current StudioBinder guide, representative of mainstream adaptation craft advice, tells writers to cut, compress, combine and write for the eye rather than the page.^[9] These are valuable tools. They become dangerous only when treated as universal laws. Sometimes the correct adaptation move is expansion.

Haruki Murakami’s short story “Drive My Car” is built largely from a few conversations and a reservoir of implication. Ryusuke Hamaguchi’s film runs for roughly three hours. He and co-writer Takamasa Oe did not discover hidden plot inside the story. They created new dramatic architecture around its absences, drawing on other Murakami material and on Chekhov’s Uncle Vanya. Hamaguchi has said that he and Oe reread the story repeatedly, not because it could simply be transferred, but because they wanted its mysteries to become internalized enough to survive extensive invention.^[10]

This is adaptation as absorption rather than extraction.

The screenwriter reads until the source stops being a list of scenes and becomes a set of instincts. Only then is radical change less likely to become arbitrary.

Lee Chang-dong’s Burning, also derived from Murakami, offers another version of the process. Murakami’s “Barn Burning” is sparse, opaque and suspended around a mystery that may not have a solution. Lee and co-writer Oh Jung-mi turn the narrator into a young man marked by economic precarity and family fracture, embedding the mystery in contemporary South Korean class tension. BFI criticism notes that the film does much more than relocate the story and reads its class anger as a distinctly contemporary transformation of the source.^[11]

The point is not that every adaptation needs to become overtly political. The point is that context is never neutral. If you move an old story into the present but fail to ask what money, gender, technology, race, class, law, labor, sexuality or public reputation now do to the characters, you have changed the calendar without changing the story’s physics.

This is why “timeless” can be a dangerous compliment. Stories do not remain alive because they float above history. They remain alive because successive artists discover new historical conditions under which their conflicts become visible again.

Re-Authoring Point of View

The same principle applies to point of view. Novels possess narrators. Films possess cameras, microphones, performers, cuts and frames. Those are not equivalent instruments. A first-person sentence can tell us exactly what someone claims to think. A close-up cannot. The close-up gives us a face that may corroborate, undermine or refuse interpretation. The loss of direct access to consciousness is not merely a deficit. It is one of cinema’s great opportunities.

Joe Wright’s Atonement is a useful case because the source novel’s moral architecture depends on perception, misinterpretation and retrospective authorship. Scholarship on the adaptation has shown how the film reorganizes point of view rather than simply reproducing Ian McEwan’s narration, using cinema’s sequencing of perspectives to make the viewer participate differently in error.^[12]. A screenwriter should therefore treat point of view as something to be re-authored from zero.

Ask not, “Who narrates the book?” Ask, “Whose misunderstanding gives the film its shape?” Ask what the audience is allowed to know, when they are allowed to know it, and what image will later become contaminated by new information. Ask whose absence organizes the scene. Ask whether the camera is intimate, accusatory, indifferent or trapped.

Jonathan Glazer’s The Zone of Interest is an extreme example of this re-authoring. Glazer began with Martin Amis’s novel, then moved away from it toward archival research on Rudolf and Hedwig Höss. He has said that the research took the project away from the novel and into primary sources, where the ordinary aspirational texture of the Höss family became central to his conception.^[13] The film discards much of Amis’s plot machinery and makes an appalling formal wager: domestic normality remains visible while industrialized murder is pushed largely into sound and peripheral evidence. BFI critic Nicolas Rapold notes that the film keeps the family’s self-absorption foregrounded while Auschwitz persists as an acoustic and spatial reality beyond the wall.^[14]

What has been adapted there? Not incident in any straightforward sense. The adaptation targets a moral condition: the capacity to build ordinary life beside atrocity. That is a far more severe test of fidelity than whether the scenes match.

The practical lesson is that internality should not simply be “externalized” through explanatory dialogue. That advice is often repeated because it is easy to teach. It also produces terrible scenes. A character who thinks beautifully in prose should not automatically become a character who speaks beautifully in film. Thought can become framing. It can become withholding. It can become an object handled too carefully, a sound heard from another room, a repeated gesture, a refusal to answer, an edit that joins two incompatible realities.

Alex Garland described his Annihilation not as a literal transfer of Jeff VanderMeer’s novel but as something like a dream of the book.^[15] That is more than a nice phrase. Dreams are not summaries. They retain emotional geometry while corrupting sequence, scale, identity and causality. For difficult fiction, “dream of the book” may be a stronger working method than “outline of the book.”

It is also a useful defense against the worst kind of adaptation reverence: the screenplay that behaves as though every scene were a protected landmark.

A Hierarchy of Fidelity

Before drafting, identify what the source can afford to lose. Then identify what it cannot survive losing. Those lists should not be the same.

A story may survive without its famous subplot but collapse if its moral ambiguity is clarified. It may survive without its original ending but collapse if its protagonist becomes more likable. It may survive a new country, century, gender configuration or genre, but die the moment a frightening contradiction is explained away.

The adapter therefore needs a hierarchy of fidelity. Not fidelity to everything, which is impossible, but ranked obligations. Perhaps the first obligation is to a contradiction. The second is to a relationship. The third is to a structure of revelation. The fourth is to a tone. Plot may come fifth.

Or the order may be completely different. Greta Gerwig’s Little Women is instructive because it complicates any crude argument for radical departure. Gerwig kept a remarkable amount of Alcott’s language, while also restructuring the novel’s chronology, beginning with the sisters as adults and letting girlhood appear as something already lost. She has explained that starting in adulthood allowed childhood to become an object of yearning shared by characters and audience; she also folded material from Alcott’s life and other writing into the screenplay.^[16]

Here fidelity to language and infidelity to structure cooperate. That matters because “fidelity versus freedom” is a false binary. The more precise question is: fidelity to what? Sometimes exact language is the danger. Sometimes exact language is the museum glass.

The new structure of Little Women changes the meaning of familiar scenes because memory now shadows them. We are not simply watching childhood unfold. We are watching adulthood remember what it cannot keep. The adaptation exploits the audience’s prior knowledge of the material. Recognition becomes dramatic material rather than a problem to overcome.

This is especially important with canonical texts. The audience may know what happens. Suspense cannot depend only on outcome. The adapter must create uncertainty elsewhere: interpretation, motive, timing, identification, tone, moral alignment.

Audience Memory as Screenplay Machinery

Psychology helps explain why variation can matter. Research on narrative transportation describes immersion as a concentration of attention, emotion and imagery inside a story world, not merely comprehension of plot.^[17] Research on prediction error and memory, meanwhile, shows that violated expectations can destabilize ongoing representations and promote memory updating.^[18] These findings do not provide a formula for art, but they clarify a useful intuition: familiarity creates predictions, and predictions create something an adapter can break.

The audience’s memory of the book is not baggage. It is part of the screenplay’s available machinery.

A famous line can be delayed. A known death can be reframed. A villain can enter through another character’s affection. A scene can begin exactly as expected and then refuse its inherited function. A canonical romance can be made to feel less inevitable by changing whose desire controls the frame. This is how an adaptation can generate danger even when the plot is public knowledge.

Compression Without Losing Complexity

The same principle should govern character compression. Screenwriters are often told to merge minor characters because two hours cannot accommodate a novel’s population. True, but the sophisticated question is not “Who can be removed?” It is “What dramatic function would disappear with them?”

If three characters each challenge the protagonist in distinct ways, merging them may reduce not just page count but moral complexity. If five characters merely deliver variations of the same pressure, a composite may intensify rather than diminish the story. Compression should increase force per scene, not merely reduce runtime.

Sarah Polley’s process on Women Talking offers a useful corrective to mechanical streamlining. While adapting Miriam Toews’s ensemble novel, Polley has said she wrote drafts from the point of view of individual characters so that each remained fully alive before the screenplay settled on its final narrative strategy.^[19] The technique is almost perversely expansive for a writer facing compression, but that is why it works. She did not reduce the group before understanding the pressure each person contributed. The efficient move came after the deep one. That sequence matters. Many bad adaptations become efficient too early.

They identify plot before theme, theme before contradiction, protagonist before ecology. They simplify the work before they know what complexity is doing. Once that happens, no amount of elegant screenwriting can restore the lost pressure.

A Practical Workflow for Adapting Literature

For a screenwriter adapting literature, a more rigorous workflow begins with diagnosis.

First, write a one-page account of the source that contains no plot summary. Describe only what it makes the audience feel, what it refuses to resolve, and what power structures determine behavior.

Second, list the five scenes readers are most likely to consider indispensable. Beside each, write the dramatic function. Then design at least one alternative scene that could perform the same function without copying the event.

Third, map the source’s interiority. Mark every important development that occurs as thought, memory, narration or interpretation rather than observable action. Do not immediately convert these into dialogue. Assign each one a cinematic channel: behavior, image, sound, performance, duration, framing, editing, production design or silence.

Fourth, draw the power map. Who can punish whom? Who controls money, sex, information, mobility, reputation, work, family, law or violence? If the setting is changing, rebuild those powers for the new context before updating any surface detail.

Fifth, choose one deliberate betrayal. Change something important enough that the adaptation must justify itself. Point of view. Chronology. Setting. Protagonist. Ending. Genre. If every major element feels too sacred to alter, the screenplay is probably still subordinate to the book.

Sixth, preserve one non-negotiable source tension. Not a scene. A tension. Love and resentment. Freedom and corruption. Desire and shame. Belief and evidence. Intimacy and surveillance. Whatever it is, this becomes the adaptation’s internal compass.

Seventh, design for the reader who knows the book and the viewer who does not. The first needs surprise without contempt for recognition. The second needs a complete film without homework. The screenplay must function simultaneously as return and first encounter.

One way to make those decisions operational is to perform a four-way scene audit before outlining the film. Mark a source scene KEEP only when its event, function and effect are still inseparable. Mark it TRANSFORM when the function matters but the original event has become dramatically weak, too interior or culturally inert. Mark it CUT when it exists mainly because prose can sustain digression, commentary or repetition that the film does not need. Mark INVENT wherever the screenplay has lost an effect that the novel achieved through voice, syntax, access to thought or duration.

The fourth category is the one adapters underuse. Invention feels suspicious because it produces material with no obvious textual alibi. Yet without invention, cinema is often asked to compensate for what prose can do by doing a thinner version of prose. A new scene can be more faithful to a novel’s emotional logic than an illustrated old one. The standard should be functional proof: what missing pressure does this invented material restore?

This also gives revision a measurable question. For every major departure, write the source function it replaces. If no function can be named, the change may be ornamental. For every major element retained, write the effect it still produces on screen. If the answer is only “because readers expect it,” the scene may be fan service rather than drama.

This method is harder than literal fidelity because it requires the adapter to be accountable for every departure.

Danger Versus Reassurance

It also guards against the opposite failure: modernization as self-congratulation.

Changing a classic is not automatically bold. Sometimes the most predictable thing an adapter can do is congratulate the present for being more enlightened than the past. Characters are given contemporary language for ideas they could not plausibly articulate. Social contradictions are flattened into approved positions. Ambiguities that made the source disturbing are corrected. The adaptation becomes less an encounter with another mind than a performance of our superiority to it.

That is not danger. It is reassurance.

The strongest modernization allows friction between eras. It does not ask the old text to pass a contemporary exam. It asks what the old text can still accuse us of, and what we can now see that it could not.

This is where fidelity returns through the back door, but in a better form.

Dudley Andrew, writing about Bazin, notes that fidelity may be overvalued without being meaningless. The source has a historical and aesthetic specificity that adaptation should not erase merely because theory has made “faithfulness” unfashionable.^[20]

There are things an adapter should be afraid to lose.

Cultural specificity is one. Formal strangeness is another. So is an author’s refusal to explain. The right to reinterpret is not the same as the right to homogenize.

This is why the most durable adaptations often feel simultaneously disrespectful and deeply attentive. They violate surfaces because they have identified deeper obligations.

Throne of Blood does not preserve Macbeth by preserving Shakespeare’s words, just as Clueless does not preserve Emma by preserving Regency England. The Handmaiden does not preserve Fingersmith by preserving Victorian Britain, and Burning does not preserve Murakami by keeping narrative minimalism at the same scale. Drive My Car does not preserve a short story by staying short. The Zone of Interest does not preserve Amis by preserving his plot, and Little Women does not preserve Alcott by preserving chronology. What these films preserve, in different ways and with different degrees of success, are systems of pressure.

That is the standard worth stealing.

It is also the useful lens for a Broken English project such as Some Of Us Refuse. The productive question is not which inherited line, episode or surface must remain untouched. It is where the refusal lives. Is it an act of agency, an emotional wound, a resistance to passive acceptance, a contradiction between private memory and public story? Once that pressure is identified, interpretation and reinvention can become forms of preservation. Chronology, point of view, setting or incident may change precisely so the underlying resistance does not become decorative.

That is not an advertisement for infidelity. It is an argument for responsibility.

A screenwriter adapting an old text should be able to answer three questions before writing FADE IN.

What in this source was once dangerous? What has happened in culture, storytelling and audience memory that has made that danger easier to ignore? What must the film change so that the danger is felt rather than merely recognized? If the answers are specific, the adaptation has a reason to exist.

If the answers are “the plot,” “nothing” and “as little as possible,” it may have only a reason to be filmed. The difference is the whole art.

Broken English stuff:

Why the Best Plot Twists Do Not Reveal New Information, They Reframe Everything. What’s all this chatter about “romantasy”? The Camera Is the Villain: Voyeurism and the Ethics of Looking The Allure of Supernatural Academies in modern young adult literature AI Beta Readers: How Machine Intelligence is Transforming Story Development Must writers now collaborate with AI to stay ahead?