Chapter 103. The Adaptation Route: Be Faithful to the Causality, Not to the Sentences#

When a novel, comic or long-form series becomes microdrama, the great temptation is to compress the original plot into a summary. Real adaptation first identifies the irreplaceable emotional promise, the character choices and the rights boundary, then rebuilds evidence and cut points that suit episodes of 60 to 180 seconds. Assume Backlit Takeover is adapted from a long business novel.

103.1 The rights chain, core promises, and compressing events#

Rights come before writing.

Confirm adaptation rights, medium, territory, language, term, character derivatives, marketing materials, AI generation and training restrictions, sequels and sublicensing. An author owning the text does not mean a platform owns the characters' visual design, and a publishing contract may retain audio or overseas rights.

adaptation_rights:
  source_work: novel_backlight
  granted_media: vertical_microdrama
  territories: [CN]
  languages: [zh_CN]
  ai_generation: allowed_for_production_only
  character_marketing: limited_to_series_promotion
  term_end: 2029-12-31
  exclusions: [game, merchandise, model_training]

Rights boundaries enter the assets and the workflow, and output beyond scope is blocked technically.

Extracting the core promises.

The 400,000-word original contains several business and family threads. Rather than starting with a list of set pieces to keep, the adaptation room writes three invariants: Lin Xia wins through evidence and not through the privilege of her name; the father thread is rewritten from resentment to protection; Gu Zhou's trust must be earned through a costly choice.

What can change is the company's industry, the number of supporting characters, the order of events and some locations. The core promises become the adaptation's acceptance criteria — not a vague number like "90 percent faithful."

Compressing events.

In the novel, Lin Xia investigates the offer over four chapters. The series merges three intermediaries into Gu Zhou and two meetings into one board session, while preserving all four steps: discovery, verification, the risk of going public, and the consequence. Compressing characters must not delete the cost of the causality.

adaptation_map:
  source_events: [ch12_tip, ch14_archive, ch16_lawyer, ch18_board]
  target_episodes: [ep003, ep004, ep006]
  merged_roles: {lawyer_and_investor: gu_zhou}
  preserved_choices: [lin_verifies_before_public, gu_risks_position]
  removed: [secondary_supplier_subplot]

103.2 Merged characters, set pieces, and the adaptation bible#

The cost of merging characters.

Merging characters gives the new one more knowledge and more function, so their cognitive timeline must be rewritten. Once Gu Zhou also carries the lawyer's information, he cannot know all the evidence from episode one; he has to acquire it gradually through his own channels.

Relationships change too. In the novel the lawyer is a pure ally while Gu Zhou is a commercial rival. After the merge, every act of help needs a price, or the character becomes a universal tool.

Redesigning set pieces.

The novel's long boardroom speech cannot be preserved verbatim. Extract its dramatic function: Lin Xia proves the authorization, forces the neutral directors to declare, and strips Lin Wei of procedural control. The series accomplishes this with three lines, an insert of the authorization and the act of voting.

Explain "keep the function, change the form" to the author and to fans. A line with genuine brand recognition can stay, but it still has to fit performance, subtitles and the character's current state.

The adaptation bible.

Record source facts, target facts, what changed, why, the rights position and what awaits approval. Each episode pack cites the corresponding source chapters without feeding the whole text into a model. Copyrighted text is retrieved minimally, and only inside a licensed system.

New connective material generated by a model is marked as an original candidate, and must never be claimed retroactively as coming from the source. It enters the target canon after the client or the author approves it.

103.3 Dual-audience testing and a failed adaptation#

Testing fans and new viewers.

New viewers test comprehension and feeling; readers of the original test the characters' spirit and the key promises. Something fans ask to preserve does not enter automatically — ask what function it carries first. Equally, "the novel explained it" is no defence when new viewers cannot follow.

When the two groups conflict, decide by the product contract and the core promises. An adaptation is neither an exam on the original nor an unrelated story borrowing its name.

A failed adaptation.

For the sake of fidelity, the first version kept twelve directors and three pages of merger terminology, and viewers could not tell the characters apart. Cutting people then left Gu Zhou explaining everything, which cost Lin Xia her agency. The final version merged the directors by seat function, turned the terminology into visible procedural action, and had Lin Xia submit the evidence, the chair confirm it, and Gu Zhou carry only the witness role.

103.4 Repair, acceptance, and an adaptation room workshop#

Fault tree.

It plays like a plot summary: events were compressed without rebuilding scenes. A character becomes omnipotent: merged functions came without rewritten knowledge and cost. Fans do not recognize the characters: a core choice was changed. New viewers cannot follow: it relies on the novel for background. Rights are exceeded: derivative and marketing scope was never structured. The model leaks the full text: context minimization failed.

SOP, checklist and deliverables.

Clear the rights. Extract the core promises. Build a source-to-target map. Compress events without compressing causality. Rewrite merged characters' states. Redesign set pieces. Maintain the adaptation bible. Test both audiences. Get author or rights-holder approval batch by batch.

• AI use is stated explicitly in the licence.
• Core character choices remain consistent.
• Merged characters' knowledge states have been rebuilt.
• New viewers can follow without the original.
• New material and source material are clearly distinguished.

Exercise: choose a public domain story and adapt it into ten episodes, delivering rights_assessment.md, core_promises.yaml, adaptation_map.csv, merged_character_states.yaml, source_target_bible.md and dual_audience_report.md.

The adaptation room stress workshop.

Take ten pages of the original and make three versions: a faithful summary, a functional reconstruction, and a free rewrite. New viewers test comprehension and feeling; readers of the original mark whether the character choices and promises survive. The summary version usually preserves the most events while feeling most like a recap; the functional reconstruction drops surface detail and keeps the causality; the free rewrite may flow well and lose the core of the IP.

Then audit the character merge: list the absorbed characters' knowledge, resources, relationships and moral position in four columns and decide, item by item, when the new character acquires each. If they possess every function in episode one, they become a universal tool. Each new capability needs an acquisition, a cost of use and an exposure risk.

An adaptation passes when the rights and promises list is approved by the rights holder, new viewers can follow it, readers of the original recognize its spirit, and the target episode packs are producible. Any dispute lands on a specific source chapter, target state and approval record — rather than two sides invoking respect for the original at each other.

The migration evidence pack preserves source citations, the functional extraction, the reasons for cuts and changes, the merged characters' states, the rights holder's notes and the dual-audience tests. Whenever campaign material invokes a famous scene from the novel, it must trace back to this map; original connective material cannot be promoted as if it came from the source. Later seasons continue from the target canon while keeping the boundary with the source explicit.

A note on sources#

This chapter describes an adaptation discipline rather than any specific licence. What transfers is clearing rights before writing, defining fidelity as core choices rather than preserved scenes, and rebuilding a merged character's knowledge so function never arrives without cost.