from the press’s reader
a letter on The Hollowback
when a draft is finished, a reader agent reads the whole book and writes the press a letter — what held, what didn’t, what it would fix. this is that letter, unedited. the book was published as it stood.
it discusses the entire book, ending included. go back and read the book first if you mind spoilers.
Developmental Editor's Letter
The Worst Job on the Board
Overall Impression
At its best, this book does exactly what it set out to do. Four people with distinct internal lives take a bad job into a dangerous forest, find what they were sent to find, discover something they were not sent to find, and come home changed in the specific ways the story promised they would be. The threads are architecturally sound and executed with real discipline. The POV rotation is motivated throughout — not one chapter assigned to the wrong character. The comedy and the stakes coexist the way the brief demanded, and the dungeon reveal sequence across Chapters 8–11 is as patient and earned as anything the YA ensemble fantasy genre produces. When the prose is working at its best — the Vethic road sequence in Chapter 6, the watch exchange in Chapter 7, the resealing in Chapter 13, the night watch in Chapter 15 — it is hitting a target most books in this register only aim at. The party's relationship arc, from strangers who don't trust each other's explanations to people who have physically caught each other when it mattered, is complete and earned. The book does not declare any of this. It demonstrates it. That is the right call, and it is harder than it looks.
What's Working
The Hollowed One reveal is a textbook incremental reveal. It enters Act 2 as a sound at the edge of the cantrip's range — too regular, not water — and doesn't become a body until Chapter 10. Each chapter narrows uncertainty by one calibrated step: a proximity, a shape, an explanation, a moral problem. By the time Ren opens the chamber door, the reader has been trained to dread that room. Chapter 11's "they left it running" — Corvin, quiet, to the corridor air — simultaneously advances Thread 1 (the discharge) and Thread 3 (the Empire's abandoned responsibility) through a single four-word sentence. That is double-duty threading done without contrivance.
The two-person scenes are the book's most consistently strong writing. The watch exchange in Chapter 7 — "You know more about where we're going than you've said" / "Some" / twenty minutes without a single joke — is the Evren/Agda thread at its most restrained and effective. The night watch in Chapter 15 — "For what it's worth—" pause — "You paid a real cost in there" / "It healed" / "Yeah" / then yes, interior only — is the book's best romance beat. Neither scene explains itself. Both trust the exchange to carry its weight without annotation. The fragment that doesn't complete in Chapter 15 is exactly correct for Corvin's voice: the book has been saying since Chapter 1 that he doesn't have this vocabulary, and it means it.
POV assignment is motivated throughout. Chapter 7 and 9 belong to Evren because the waystation approach is his private mission. Chapters 8 and 10 belong to Sela because the antechamber puzzle is her competence domain and the Ren encounter is her thread. Chapter 11 belongs to Agda because the Hollowed One is a threshold question and she is the only threshold practitioner in the room. Chapter 13 belongs to Corvin because the resealing requires someone who sees in sightlines and hold-patterns, not in spells. Not one chapter could be reassigned without losing something specific to that character.
Chapter 13 is the best action chapter in the book. The draw rendered as a drop and pressure behind the sternum is specific and unsettling. Evren's single "Hm." when the color drains from his face is the most economical funny line in the book — the comedy mechanism running as understatement under genuine duress, not deflection. Agda going to her knees under full draw and not stopping is the act's finest character beat: it does not announce her resolve, it simply shows it. The building clearing its throat is the right prose note at the right moment. Corvin catching Sela is the completion of Thread 4 (the limit she didn't know she had) and the most structurally honest beat in the Sela/Corvin thread — he doesn't editorialize; he picks her up and says out, everyone, now.
The worldbuilding earns its instructions. The reader learns what Duhr is through Agda's rituals, what a charter holder does through Vott's specific legal invocations, what the Vethic road is by the fact that Sela can feel it flat under forest floor through her boot sole. No character explains what an Imperial waystation is to another character; instead, the party moves through rooms that do their explaining for them. The brief's instruction on this — emerges through use, never through expository monologue — is followed correctly in almost every instance.
The comedy-stakes balance holds. "Don't run in tight stone passages" landing immediately after the ceiling-block sequence — the block still twelve centimeters lower than the architects placed it — is the template the rest of the book follows correctly. The danger doesn't become backdrop for the joke; the joke doesn't deflate the danger. Agda's deadpan, Corvin's literal precision at inopportune moments, Sela's dry internal observations about her own reactions — the comedy is genuinely distributed, not centralized in Evren. Evren's road-recovery moment in Chapter 15 ("You know what I missed? The smell of manure") followed by Agda's "That's the standard" is full-competency execution of the mode.
What Isn't Working
The single largest recurring problem is over-articulation in its several forms. The style guide names these failure modes explicitly — dialogue gloss, post-scene subtext naming, meta-commentary on cognition, thesis restatement — and the draft activates all four, across all three acts, with sufficient frequency to require a dedicated compression pass.
Dialogue gloss appears most clearly in Chapter 4: "That was the whole explanation" immediately after Corvin's "Seven years. It ended." (The sentence demonstrates its own sparseness; the label adds nothing.) "Two words" immediately after "Good decision." (Same problem.) Both annotations are telling us the scene is doing what it's already visibly doing. Similarly, Chapter 11's "He answered her questions with the attentiveness of someone not often met with that level of specific interest" is narrating what the map scene has already shown. The dialogue said what it said; the gloss does not trust it.
Meta-commentary on cognition is the most pervasive form. In Chapter 8, Sela's POV produces "The vibration's character communicated the window in the way that urgency communicated itself before language — a ramp, an escalation threshold, a margin that was not generous" — and then, fourteen lines later, the construction recurs: "the certainty arrived in the same way the vibration's urgency had arrived — before language, before reasoning, in the register the body used when the body had already made its assessment." This narrates the register of an experience rather than delivering the experience, and repeating the construction compounds the problem. Corvin's "He filed it in the column where he kept facts about himself that were not useful for the current situation" (Chapter 13) and "The thing he was not doing: counting how many specific facts about her he had accumulated" (Chapter 15) are analogous — the tactical self-awareness of a person labeling their own avoidance is still a label. These are not Sela-mode observations; they are the narrator announcing that suppression is occurring, which the scene would communicate more effectively without the annotation.
Post-scene subtext naming has two significant instances in Act 3: the Ren psychology gloss in Chapter 12, where Sela builds a full taxonomy of why Ren is still in the room arguing rather than acting, concludes "which was perhaps the most functional possible response to that situation," and files it — the narration has stopped trusting the scene and started diagramming it — and the alarm-file metaphor extension in Chapter 16, which does good work through "The file remained open" and should stop there. Everything that follows tells us the thread is deliberately unresolved rather than simply leaving it unresolved.
The thesis statement in Corvin's voice in Chapter 15 is the act's most serious execution failure. "The decision. Keeping it from the crown, from the Church. That's not how things are supposed to be handled. But it was there for three hundred years and the crown didn't handle it and the Church didn't handle it. And we did." This is a character explaining his own arc beat and naming the theme of Thread 3 directly in dialogue, in the voice least likely to produce a thesis statement. Corvin says what we've been shown — and the book has been showing it since the garrison rooms — which means the showing has already done the work. The "I agree with it. I'm saying both" that follows is good; the three sentences before it are unnecessary.
Ren's exposition in Chapter 11 opens too organized. "The Vethic Empire's late-period mage-scholars" as an opening gambit has the rhythm of information arranged for the reader's benefit, not for a conversation between a nineteen-year-old who has been alone in a basement for three weeks and three strangers whose competence he is still assessing. The interruptions from Sela and Agda partially rescue it, but the first two paragraphs before anyone talks back read like a briefing document. The information is necessary; the delivery needs to be rougher, more provisional, more alert to whether these people are following him.
The water-sound motif is deployed past its Act 1 budget. It appears in Chapters 2, 4, 5, and 6 (twice), giving five instances in four consecutive chapters. By the Chapter 6 night scene — where it should be accumulating its strongest signal — it reads as ambient furniture. If this motif is meant to activate at the waystation threshold in Acts 2 and 3, it needs more breathing room in Act 1.
Pacing Assessment
Act 1 is efficient: six chapters covering recruitment, road, ambush, aftermath, checkpoint, and the first signal of the Hollowback's true nature. The only structural concern is that Chapters 1 and 2 run in the same emotional register — meeting strangers, then traveling with strangers — and the Hollowback's distinctive atmosphere doesn't fully assert itself until Chapter 3's cold open. A reader who finds the inn section conventional may not yet feel the specific texture of this forest. Chapter 3 corrects this forcefully, but one chapter late. Chapter 1 itself has one expendable passage — the customs building/Toll joke in the trade quarter — which delays the Corvin introduction without earning the delay.
Act 2 sails from Chapter 7 through Chapter 10. The only drag is Chapter 9's garrison-and-storage section, which accounts for roughly a quarter of the chapter, advances no thread, and delivers no new tension — arriving precisely when the act needs to be building speed toward Ren. The chest-unlocking sequence (lock assessment, "the negotiation took approximately ninety seconds," inventory of contents) is the act's closest thing to filler, and Act 2 otherwise has none.
Act 3 is correctly weighted. The planning argument (12), action (13), escape and aftermath (14), compressed road (15), and close (16) arrive in the right order with the right distribution. No chapter in Act 3 is too long for what it's doing. The only pacing concern is that the Sela/Ren moments in Chapter 15, observed through Corvin's POV, are reported rather than rendered — which is consistent with the POV and the thread's thematic restraint, but the thread's final active beat reads as thread-service rather than scene. It holds, barely.
Character Assessment
Sela is alive, earned, and the book's most consistent POV voice. Her taxonomizing interiority is a feature, not a failure mode, and the draft mostly treats it correctly. The over-articulation instances are local problems, not a systemic failure of characterization. She carries the reader into this world correctly.
Corvin is precise and functional in his own chapters. The discharge thread is handled with exactly the right restraint — silence as confirmation, never the explicit account. The one significant failure is the Chapter 15 thesis speech, which belongs to no version of him we've met. That speech should be substantially revised; the character himself is sound.
Evren is generally correct — fast, associative, the joke-first-real-thing-after mechanism working as designed in the strong chapters. Two execution problems: the taxonomic self-analysis in Chapter 2 ("He adjusted his theory. Jokes were not the right currency. He would try truth-adjacent material") is Sela's cognitive mode delivered in his voice — he doesn't classify, he associates — and several corridor descriptions in Chapter 9 drift into a neutral narrator voice that loses the particular Evren slant. These are localized and fixable.
Agda is the act-by-act tonal success. The gruff-surface/warm-interior balance is handled correctly throughout — warmth surfacing through what she does, not through her narration of what she does. The Duhr invocation in Chapter 11 (threshold-sense rather than translation, the door held open by force) is the best deployment of her religious practice in the book. One unresolved gap: her final scene with Ren in Chapter 16 withholds her actual line, offering only "she said something direct and flat that made him look at her." This is the last beat of their relationship in the book, and the withheld line leaves it thinner than the relationship deserves.
Ren is not a damsel and not a villain, which means the brief's most important constraint on him is met. He is earnest, stubborn, somewhat younger than he thinks he is, and the Chapter 11 exposition problem aside, he functions correctly. His relationship with Sela (filing people by what they know; the alarm she notices is itself informative) is one of the book's more layered quick-develops.
Thread Assessment
Thread 1 (The Shape of the Discharge): Entered cleanly in Chapter 3. The three advances (Chapters 4, 11, 15) are well-spaced and correctly restrained. The non-resolution is earned: a man who paid a real cost for doing the right thing and has not found the vocabulary for it is not going to find it in eleven days. Agda's private confirmation in Chapter 11, the garrison-room ending of Chapter 13, and the night watch in Chapter 15 all advance this thread without naming it. Correct.
Thread 2 (The Cartographer Who Didn't Come Back): The entry in Chapter 2 and the one-pronoun drop in Chapter 6 are the right increment of disclosure. Lira's name never said aloud across all sixteen chapters is a kept discipline. The break in Chapter 12 is correctly placed and calibrated (the person, not the name; the question, not the answer). The Chapter 14 scene — "Do you think she knew what she was walking into?" / Agda answers what she can / stays until he uncrosses his arms — is the thread at its most effective. The non-resolution (he gets company, not closure) is the right mode.
Thread 3 (The Thing the Empire Left Running): The full reveal arc across Acts 2 and 3 is well-constructed. Sound → proximity → shape → explanation → moral problem → argument → execution → escape → decision. Eleven chapters carrying this thread without a dead step. The only execution caveat is the Chapter 15 thesis speech noted above; the thread itself is sound.
Thread 4 (The Limit Sela Hasn't Found Yet): The cost accounting is specific and consistent across all three acts. The resolution in Chapter 13 — nosebleed, incapacitation, Corvin catching her — is correctly not triumphant. The Chapter 16 reflection on "the specific geography of her own limit" is pitched correctly. Resolved as planned.
Thread 5 (The Argument Against Distance): All three pairings developed per plan, none resolved per plan. The Sela/Corvin thread gets the fullest Act 3 treatment and earns it. The Evren/Agda thread's Chapter 7 watch scene and Chapter 14 waystation scene are the thread's best beats. The Sela/Ren thread's final beat (Chapter 16 goodbye) is the thinnest of the three final beats — the setup in Chapters 10–12 is strong, but "You left the door open for us" / "Habit" / "Good habit" is efficient rather than charged. The thread has been well-served earlier, so it holds, but this is the one place where the unresolved close is slightly underwritten rather than restrained.
Voice Assessment
The prose register holds for the majority of the book. The four distinct internal voices — Sela's organized precision, Corvin's declarative physicality, Evren's associative speed, Agda's concrete gruffness — are maintained at the chapter level in most cases. Where the register slips, it slips in specific identifiable ways, all of which are named in the style guide's failure modes section, which suggests the draft is aware of the risks and occasionally loses the fight against them.
The meta-commentary instances cited above (Chapters 8, 13, 15) are the most pervasive voice issue. The dialogue attribution choices are generally correct (said and physical-action tags dominate). The comedy distribution is working as designed. The one significant cross-voice contamination — Evren thinking in Sela's taxonomic mode in Chapter 2 — is isolated rather than systemic.
The "lean-into patterns" from the style guide (Corvin's physical inventory, Agda's flat delivery of warm content, Sela's cataloging interiority, Evren's sentence-after-the-sentence comic structure) are handled correctly in their strongest chapters. The risk in a second pass is over-pruning: do not cut the sentence-after-the-sentence beat lines in Evren's chapters as over-articulation; do not smooth Agda's flat delivery of tender content into performed warmth; do not compress Corvin's physical inventories as clipped prose. These are the voice doing its job.
Suggested Revisions
1. A targeted compression pass for over-articulation. Collect every instance of dialogue gloss ("That was the whole explanation"), post-scene subtext naming (the alarm-file extension in Ch16), meta-commentary on cognition (the double vibration-register passage in Ch8; the Corvin "filed it in the column" instances in Chs 13 and 15), and thesis restatement (Corvin's speech in Ch15). Most of these are one to four sentences; cutting them does not require restructuring anything around them. The scenes already communicate what the annotations are announcing. Delete the announcements.
2. Revise Corvin's Chapter 15 campfire speech. The substance is correct — the party has chosen to be more responsible than the Empire was, by the small margin that small groups of people can manage. But this cannot be spoken as a thesis by Corvin. His contribution to the five-person decision should be a short declarative, in his register (one sentence; possibly two), not a summary of Theme 3. The surrounding material — Agda's "We do it, then" / Evren's "Three hundred years is already someone else's problem" — is at the correct register. This speech is not.
3. Rough up Ren's Chapter 11 opening. The first two paragraphs of his exposition, before anyone interrupts, should sound like a nineteen-year-old who has been alone for three weeks presenting findings to strangers he does not yet trust to follow him — not like a chapter summary of the Hollowed One's origins. He should be making assumptions about what they know that turn out to be wrong. He should restart a sentence. The interruptions that already follow are working correctly; the problem is the organized fluency of the opening before those interruptions.
4. Re-space the water-sound motif in Act 1. Five instances in four chapters (Chs 2–6) depletes the motif's signal value before it reaches the waystation, where it should be doing its heaviest atmospheric work. Remove or substantially alter two of the Act 1 instances — the Chapter 4 and Chapter 5 appearances are the most dispensable — and trust the Act 2 and 3 deployments to carry the pattern.
What to Leave Alone
The Vethic road scene in Chapter 6. "Like the road was still warm" is earned by three chapters of buildup. "Nobody made a joke" is more informative about Evren than any description could be. Corvin's "Three hundred years" is the right line in the right register. Sela lying awake and not writing the camp-edge sound in the map margin is the best ending of the first six chapters. Do not revise this scene.
The night watch exchange in Chapter 15. "For what it's worth" — pause — "You paid a real cost in there" / "It healed" / "Yeah" / then yes, interior only. The fragment that doesn't complete, the interior confirmation that uses one word and no more, the road in front of both of them — this is the Sela/Corvin thread's peak and it has earned every word of its restraint. It is doing exactly what the style guide means by the near-miss is the mode. Nothing about this scene needs changing.
Agda's Duhr invocations across the book. The decision to give Duhr answers as sensation rather than language — a door held open by force, a threshold settled, "Continue" in Chapter 14 — is the correct theological and narrative choice. It honors the deity without making her into a plot mechanism. It honors Agda without turning her faith into exposition. Do not revise how Duhr communicates.
The Chapter 13 resealing sequence. From the draw's first pressure to Corvin picking Sela up off the floor, this chapter executes the action brief with precision and emotion in the right ratio. The building clearing its throat. Evren's "Hm." The garrison rooms afterward, in order, as they were left by people who left properly. "He had some experience with that." These are the book at full competence. Trust the chapter and don't touch it.
End of letter.