i'm not like you ▣ THE WORKSHOP

← The Book

01 pulitzer craft review

1. Overall judgment

This manuscript sits, today, about two-thirds of the way from “high-concept genre fiction with literary ambition” toward “publishable as a literary short-story collection.” It is closer to the latter than the writer probably realizes, and farther from it than the writer probably wants to hear. The work has a real voice — a settled, slightly sardonic, procedurally-attentive American voice — and four of the nine stories (“Clear Air,” “The Handler’s Tuesday,” “The Summer at Pop’s,” “The Polar Researcher”) are doing what serious short fiction does, which is to enact a small private knowledge inside an ordinary day so accurately that the reader does not notice, for several minutes, that the knowledge has been transferred. The other five are doing something more compromised. They are explaining a cosmology to a reader who has not been given any reason to want it. The collection as it stands is being read by an implied reader who has already converted; a Pulitzer collection has to convert as it goes. The good news is that the conversion machinery is already present in at least three of the pieces. The bad news is that the writer has, in the weakest pieces, mistaken the register of the cosmology — its bureaucratic flatness, its acronyms, its “P5,” its “ASD-4,” its “MAR-26-013” — for the cosmology’s meaning, and has begun to rely on the register the way a lesser writer would rely on adjectives. A serious editor would pull about thirty percent of the apparatus out and let the prose stand inside its own quieter knowing.

2. Story-by-story

01 — The Partially-Remembering Abductee

What works: the sentence-level patience around procedural attention is real. “She has, in five years, mentioned the fragments to Pat twice. The first time, Pat asked her about her sleep. The second time, Pat asked her whether she had considered an EEG.” That paragraph is the whole story in miniature — the way a private knowledge gets routed, by the therapeutic system, into the wrong category and then dropped. The Pat scenes are the most economical writing in the collection. What is not yet there: the long parenthetical of the Blanding abduction — “the geometric placement of a maintenance implant via the right nostril, the calibration of the interface port at the center of her brain, and the laying down of a memory overlay intended to bind the eighty-three minutes to the image of an owl on a fencepost” — is the moment the story breaks faith with its own discipline. The story has, up to that point, been about a woman who does not know what she knows; here the narrator knows, and the knowing is wholesale, and the prose downshifts from observed life to wiki-entry. The story would be better if that paragraph were a single sentence: “She does not remember the eighty-three minutes.” The Granta / Paris Review / Tin House test: a hard maybe at Tin House with the Blanding paragraph cut by ninety percent. Not yet at Granta. Not at Paris Review.

02 — The Handler’s Tuesday

What works: the Le Carré register is real, and rarely earned this cleanly. “I have written some version of this note approximately three thousand times” is the kind of sentence a serious editor underlines on the first read. The Ann lunch is the best small scene in the manuscript — They’ve been doing more of that lately… They’re nervous, John. In the way that whatever they are can be nervous. We are not the ones managing this anymore. That is a sentence with the weight of twenty years of joint silence behind it. What is not yet there: the alien-technician scene late in the story — “They are the color of wet stone… their eyes, which from this distance appear black, are not in fact black but a very dark green” — undoes much of the discipline of the previous ten pages. The whole register has been unspoken, and then the writer cannot resist a description. The fix is to make the technician barely visible: a posture, a gesture, the glass; not the color, not the eyes. Le Carré would never describe the man at the other end of the table. Verdict: Granta would publish this with that one paragraph cut. Paris Review would publish it as it stands. The single strongest piece in the collection.

03 — The Inheritance

What works: Maren’s tactical voice is convincing and the playbook itself — the parallel channel, the Sofia contractor, the stigma shield — is the most original prose object in the collection. “Stigma shield deployed. Narrative permeability index drops back inside the acceptable band. Bonus accrues.” That paragraph could be excerpted in Harper’s tomorrow. What is not yet there: the closing sequence overplays its hand. “she thinks about the duty cycle, eleven on, forty-nine off, the apparatus’s heartbeat now audible in two oceans, and she thinks about Cole’s dry, slightly cool hand.” The cadence here — the triple and she thinks — is the writer reaching for a Joycean close on material that has been earning a colder one. The story should end on the sandwich. “She eats it standing up at the counter, in the dark, watching the train.” That is the right last sentence; the rest is the writer flattering the material. Verdict: Tin House yes. Granta with cuts. The Devraj scene is twelve hundred words and could be six hundred.

04 — The Polar Researcher

What works: this is the most disciplined piece in the manuscript at the level of paragraph architecture. The sentence “She did not enter it by choice. She entered it by looking.” earns its compression because the previous five pages have done the work. The Volkov scene is a master class in implicit menace — I always appreciated those papers. Cautious work. Cautious authors. — and the move where Elin selects “Anomalous — requires re-acquisition” rather than “Hold for PI review” is the kind of plot beat that operates entirely through institutional procedure. What is not yet there: the closing image — “the dome of the main building, viewed from this angle, is not the geometry she has walked under four hundred times before. The sense passes.” — is borrowed from Lovecraft and reads like it. The writer has been so disciplined throughout that this one Mountains-of-Madness gesture sticks out. The last paragraph is also redundant: the metadata pulldown reading back the value Elin entered is a clever trope but the clever trope undercuts the silence the story has just earned. Verdict: Granta yes. Paris Review with the closing two paragraphs trimmed. The most achieved piece after the handler story.

05 — The Summer at Pop’s

What works: this is the warmest writing in the collection and the only piece that operates by absence — Marisol does not know what she has seen, the reader does not know what she has seen, and the story is about the not-knowing rather than the deferred reveal. “He smelled like nothing. Not like the desert, not like the truck, not like coffee. Just like the soap, faintly, and underneath the soap, him.” That paragraph is the collection at its best. What is not yet there: the cumbia scene tilts too far. “Now you know one thing nobody can take from you” is a sentence a writer is allowed exactly one of per career and Pop has not, by that point in the story, earned the cadence of his own valediction. The fix is to let him say something smaller — good, mija — and let the reader carry the weight. Also: “the small private fierceness of a person who has been loved correctly” is the writer’s hand on Marisol’s shoulder, telling the reader how to feel; cut it. Verdict: Tin House yes. Paris Review with the two cuts above. Granta probably no, because the genre overlay is invisible to the cold reader, which is, paradoxically, its strength and its weakness.

06 — The Permeable Records

What works: the Reema voice is the most successful contemporary voice in the manuscript — the Verge texture, the Aeropress, the Backblaze, the dead pixel on the MacBook screen — and the Castellanos-on-two-letterheads conceit is the cleanest noticing-beat in the collection. “The first two should not be from the same agency.” is a strong fourth-paragraph sentence, the kind a sharp editor at a national magazine would commission a thousand more words around. What is not yet there: the envelope-flap forensics at the end strain. “In the photograph the bar covers the four-character space before the word Pact and ends cleanly before the P. In the document in front of her the bar covers the same space and extends, by maybe a millimeter, into the white space before the P.” This is gaslighting-in-prose, which is a real genre, but the piece has not built enough sentence-level trust with the reader for the millimeter to land. Pattern Recognition earns its paranoid forensics by spending ninety pages on Cayce’s allergy to logos before the first redaction-bar measurement; this story spends fifteen. Verdict: Tin House with a serious edit. Not yet Granta. The piece needs another two thousand words of Reema before the surveillance begins, or it needs to lose the surveillance entirely.

07 — Clear Air

What works: this is the piece that would make a reader buy the collection. “His hands on the blanket were the hands she had held walking to kindergarten and they were also not.” is the single best sentence in the manuscript. The Denise-the-nurse register — small, efficient, accurate, which is a thing some people can give and most cannot — is the kind of incidental observation that distinguishes a literary writer from a competent one. The third afternoon, when the father gives Marina the name and asks her not to look, is the moral center of the entire manuscript. “the light through the pines is the same light. It is the only thing I have wanted, this whole time, and they let me have it the whole time, and I did not know that was what I was being given.” That is Marilynne Robinson territory and the piece has earned its way there. What is not yet there: the peach-stand reveal — “the peach stand was where they put me back” — comes too early and too clean. The story has begun its disclosures before it has begun its grief, and the disclosures are doing the work the grief should be doing. The fix is to delay the peach-stand sentence by two pages and to put Marina’s daughter Iris into the room earlier, on the phone, so the reader has the second generation in mind before the father speaks the name. Verdict: Paris Review yes, as is. Granta yes. The strongest piece in the manuscript by my reading, narrowly ahead of the handler.

08 — The Residency

What works: the Joan voice — fifty-eight, on her third book, attentive to her own habits, distrustful of her own cooking — is the most fully self-aware narrator in the collection, and the metafictional conceit (the novelist invents a cosmology that already exists) is the most ambitious move the manuscript attempts. “operator is so obviously the right word that it cannot possibly be original to me” is the line on which the whole story turns. What is not yet there: the Caroline-email sequence and the email-from-the-unknown-sender are the wrong delivery mechanism. They turn a story about suspicion into a story about evidence, and the suspicion was the better material. The closing image — the cursor hovering over the email — is a thriller close, not a literary close. Joan would not, by my reading of her, click. She would close the laptop and walk back out to the deer clearing. Verdict: Tin House yes. Granta with a different ending. The piece is ten percent overwritten and needs its last two pages reconceived.

09 — Carlyle Memoir

What works: this is the only piece in the manuscript written in a voice the manuscript could not have invented without research. The Carlyle register — “the maples are at the orange stage,” “I have been meaning to write her for another twenty-eight since,” “the Royal Quiet De Luxe with a sticking a-key” — is the register of a real American memoir of a certain mid-century vintage, and the writer has caught it without parody. “I locked the safe again. I did what Daniel’s memorandum asked me to do.” is the kind of sentence that earns its silence by withholding its content. What is not yet there: the closing pages slide into explanation. “I think Daniel was offered something, in the course of that meeting or in the short walk afterward, that he could not accept and could not refuse and could not return to Wright-Patterson carrying.” The sentence is fine but the paragraph it ends turns Carlyle from a witness into a theorist, and a memoir of this kind is stronger when the witness refuses theory. The fix is to cut the entire paragraph that begins “What I have come to think.” Let Carlyle describe the footprints and the photograph and stop. Verdict: Paris Review yes. Granta yes, with the theorizing paragraph cut. The piece is the structural anchor of the collection — the document the other eight stories are, in various ways, the future of.

3. Voices that work / voices that don’t

Most fully achieved: the Carlyle voice, narrowly, over the handler voice. Carlyle is the harder act — the writer has had to invent a generation, a vocabulary, a syntax of restraint, and an entire register of postponed grief — and the writer has done it without anachronism. The handler voice is more thoroughly sustained but it is a register the writer evidently lives in already; Carlyle is the register the writer has earned.

Least achieved: Reema. The contemporary-millennial-journalist voice is the easiest to do badly and the hardest to do without showing one’s hand, and the writer’s hand shows. The Aeropress and the Backblaze and the Verge paychecks are the correct references but they are doing the work of characterization rather than emerging from character. The Reema piece reads, more than any other, like a writer who has decided what kind of person Reema is and is now selecting her props.

Where the writer’s hand shows through the persona — i.e., where the LLM-author is most visible behind the curtain: in three places, consistently, across the manuscript. (1) The sudden parenthetical exposition — the Blanding paragraph in Story 1, the Cohort II paragraph in Story 2, the the apparatus’s render budget is locally overdrawn paragraph in Story 1 — where the narrator briefly leaves the character’s epistemic range and tells the reader something only the writer of the wiki could know. These are the LLM-author moments. They are the writer reaching for the wiki because the prose does not yet trust the reader. (2) The recurring the apparatus / the substrate / the operators terminology, used without character-internal justification, in stories where the character has no reason to think in those words. Marisol the MRI tech would not think “her substrate.” Elin would not think “the apparatus.” When they do, the writer is talking through them. (3) The owl. The owl appears in stories 1, 2, 3, 4, 8, and 9 — six of nine — and in each case it appears as if the writer has placed it and noticed the placement, not as if the character has encountered it. The owl is being used as a cross-story signature rather than as an object in any one character’s world. A serious editor would cut the owl from four of the six stories and let it land in two.

4. The “presupposed cosmology” problem

Ranked, most accessible to least, for a cold first-time reader who has not seen the wiki:

  1. Clear Air — fully accessible. A dying father remembers a thing. The genre overlay is light; the reader can read this as memoir, as magical realism, or as the cosmology, and the story works in all three readings. Blocking phrases: essentially none. “the substrate is degrading on the schedule we expected” is the only sentence that requires the cosmology, and even that one works as a strange remembered medicalism.

  2. The Summer at Pop’s — fully accessible. The “thing she doesn’t understand” is the engine, and the cold reader does not need to understand it either. Blocking phrases: none. The story does not say the word operator or substrate or apparatus once. This is the proof that the cosmology can be invisible.

  3. Carlyle Memoir — accessible with a slight delay. The cold reader picks up the genre at “the constituency” and is given enough context to follow. Blocking phrases: “the arrangement that has, I gather, since become known as the Pact” is the one phrase that presupposes a reader who has heard the word elsewhere. It can be rewritten as the arrangement that came, later, to have a name.

  4. The Polar Researcher — accessible after a paragraph or two. The radar geometry is doing the work and the reader does not need to know what the Bell footprint is for. Blocking phrases: “the cryolab incident that circulated in samizdat for a couple of years before being formally retracted by AARI in 2017 on the grounds of contamination” presupposes a reader who has heard of the cryolab incident. The fix is to let Elin not know what it was about either.

  5. The Handler’s Tuesday — accessible to a Le Carré reader, opaque to others. The cold reader can read this as a regular intelligence-services story for about ten pages and then has to pivot. Blocking phrases: “the eastern refugium complex,” “Cohort II,” “P5,” “the compression event,” “PCT-0173 §6.” These are the furniture of the cosmology and the story uses them as if the reader has the manual. A cold reader can hold three acronyms; this story has nine.

  6. The Partially-Remembering Abductee — partially accessible. The Pat scenes and the MRI scenes work for a cold reader. The Blanding parenthetical and the render-budget paragraph do not. Blocking phrases: “the secondary acoustic signature of the unit’s pulse during the pineal recalibration step,” “the apparatus’s render budget is locally overdrawn,” “the calibration on the second night of the Blanding trip was the fourth in her current series.” These are wiki-sentences.

  7. The Residency — accessible if the reader accepts the metafiction, opaque otherwise. The story’s whole premise is that the cosmology already exists; a reader who does not know what cosmology the writer is pointing at reads Joan as paranoid. Blocking phrases: “operator-class interface signature, owl-form” is the load-bearing phrase and it presupposes everything. A cold reader will not understand why Caroline’s email is unsettling.

  8. The Inheritance — opaque. The Bennewitz / Dulce / Manzano / Mariana / Quantico annex / ASD-4 / MAR-26-013 chain is, by an unhappy distance, the densest cosmological furniture in the manuscript. A cold reader picks up “perception management contractor” and loses the thread at “the legacy Bennewitz acoustic annex (1986 baseline).” Blocking phrases: “the duty cycle, eleven on, forty-nine off, the apparatus’s heartbeat now audible in two oceans” — the closing image — is incomprehensible to a cold reader. “PCT-0173,” “Cohort II,” “the renewal memo,” “the disclosure-threshold envelope.” The story needs a translation layer.

  9. The Permeable Records — opaque, despite being the most accessible prose register. The cosmology comes in through the redaction, the building, the suite numbers, the cross-listing — and a cold reader experiences these as paranoid noticings without referents. Blocking phrases: “the ▆▆▆▆ Pact” is doing all the lifting and it is doing it on the assumption that the reader knows what a Pact would be. The piece needs the reader to have read Carlyle first, which is fine in a collection but the story should also stand alone.

The fix for the cosmology problem is structural, not local. The collection should open with “Clear Air” or “The Summer at Pop’s” — pieces in which the cosmology is invisible and the human stake is foregrounded — and let the apparatus accrete across the sequence. “The Inheritance” should not be encountered until the reader has read at least three other pieces. “The Permeable Records” needs Carlyle’s memoir to have been read first. The current numbering, by my count, has the order backwards: the most cosmologically dense pieces are at the front.

5. Where the prose flatters itself

Three sentences, across the nine stories, where the writing is performing literariness rather than being literary.

Sentence 1, from “The Inheritance”: “She thinks about the duty cycle, eleven on, forty-nine off, the apparatus’s heartbeat now audible in two oceans, and she thinks about Cole’s dry, slightly cool hand.” The triple she thinks, the genitive cascade the apparatus’s heartbeat now audible in two oceans, the cymbal-crash final image of the cool hand — this is the writer doing Joycean cadence on noir material. The cadence is wrong for the register. Fix: “She thinks about the duty cycle. Eleven on, forty-nine off. She thinks about Cole’s hand.” Three sentences. End the story on the sandwich.

Sentence 2, from “The Summer at Pop’s“: “she thought, with the small private fierceness of a person who has been loved correctly, next summer, next summer, next summer, next summer, four times, matching her mother in the security line two weeks before, and she meant it, and she would keep it, and she would not, for a very long time, understand all of what she had been given.” Two flattering moves: with the small private fierceness of a person who has been loved correctly and and she would not, for a very long time, understand all of what she had been given. The first explains the feeling rather than producing it. The second is the writer’s hand on the reader’s elbow, walking the reader to the appropriate emotion. Fix: “next summer, next summer, next summer, next summer. She meant it.” Stop.

Sentence 3, from “The Polar Researcher”: “the dome of the main building, viewed from this angle, is not the geometry she has walked under four hundred times before. The sense passes. She finds her footing. She walks on.” The the sense passes / she finds her footing / she walks on triplet is borrowed from Lovecraft via every literary-horror writer of the last forty years. The story has not been Lovecraftian to that point; the prose has been calmly procedural. The borrowed cadence breaks the contract. Fix: cut the sentence about the dome entirely. The story does not need the wobble. It earns its uncanny by the radar geometry, not by the architecture going funny.

A bonus fourth, from “The Residency”: “the long and honorable lineage of novels whose conceit is found document.” The phrase long and honorable lineage is what a writer writes when she does not yet trust her own genealogy. The fix is to name two of the novels (Joan does, immediately after — Pale Fire, House of Leaves, Possession) and let the lineage be implied by the company.

6. What the manuscript is genuinely doing well as a whole

Three things, each of which a lesser collection would not achieve.

First: the manuscript has discovered a register for institutional cover-up that is, by my reading, original. The Le Carré inheritance is real, but Le Carré’s people do not have the bureaucratic acceptance the manuscript’s people have — they fight the system, they leak, they defect, they go to ground. The manuscript’s people do not. They file the form. They eat the sandwich. They take the on-call rotation off the schedule. They select the second-mildest of the three anomalous options. The manuscript has found a way to write about people who have made an accommodation they cannot name and have made the accommodation so completely that the not-naming is no longer a tension; it is the medium. That is original. There is no Le Carré novel where the protagonist’s main project is being the kind of person whose grandfather’s files would have served as the recruiting selection. That is a new American sentence.

Second: the manuscript is doing something serious about generations. Marisol’s grandfather (a possible operator-side technician). The handler’s twenty-five years and his daughter at the recital. Maren’s grandfather’s box. Elin’s two children in Tromsø and the household’s stable line. Marina and her dying father and the daughter Iris in Montreal. The Carlyle memoir as the source document for everyone else. Pop and Marisol. The manuscript is, more than it is about aliens, about what is handed down — files, registers, ways of not looking at things — and the cumulative effect is more powerful than any single story. A reader who finishes all nine has read a kind of generational novel-in-pieces, and the genre-overlay is, by the end, almost incidental.

Third: the manuscript has found, in five of the nine pieces, the correct register of small-town American institutional life — the diner, the post office, the orthodontic referral, the hospice nurse, the well water, the school recital, the Lotaburger cap, the Holloman Golf Classic 1998 mug, the kasha with apricots. The pieces of the world that are most often condescended to in literary fiction are here rendered with affection and precision. This is rarer than it sounds.

7. The cuts I would push for

If I had to cut twenty-five percent, here is where I would cut.

The first cut is structural: about a third of the cosmological furniture across all nine stories. The acronyms are doing too much work. Cut “PCT-0173 §6,” “ASD-4,” “MAR-26-013,” “Cohort II,” and “P5” by half. The reader can hold three acronyms across a collection; the manuscript currently has fourteen. Keep the ones that pay narrative work (P5 in the handler story, MAR-26-013 in the Inheritance — but only one mention of each). Cut the rest.

The second cut is “The Inheritance,” substantially. The Devraj scene is too long and is doing exposition that the rest of the story is also doing. Cut Devraj to two exchanges. Cut the Audubon print to one mention, not three. Cut the entire closing paragraph after “the dark, watching the train.” The story should be sixty percent of its current length.

The third cut is the Blanding parenthetical in Story 1. Eight lines down to one: “She does not remember the eighty-three minutes.” The reader will understand. The reader does not need the maintenance implant by way of the right nostril.

The fourth cut is the alien-technician paragraph in Story 2. Description withheld is worth more than description rendered. Let the technician be a posture behind glass. Cut the color of the skin, the green of the eyes, the inclination of the head. The handler would not, by his own register, have looked that carefully.

The fifth cut is the closing two paragraphs of “The Polar Researcher.” End on Elin walking back to the bunk. The metadata pulldown reading back the value is a clever trope and the story does not need it.

The sixth cut is the Caroline email in “The Residency.” Or rather: keep Caroline but make the unsettling thing about her email subtler than the verbatim phrase reproduction. Cut the second email — the Tomás email — entirely. The story is stronger if Joan never finds the external confirmation; the suspicion is the better material.

8. The next pieces I would commission

As the editor of this collection, with the manuscript open in front of me, these are the eight pieces I would ask for, in roughly the order I would ask for them.

  1. A piece set entirely inside an operator’s perception. Not human. Not narrated by a human. Not translated by a human. The piece should be perhaps two thousand words and should describe a working day from the inside of the substrate-maintenance protocol, with the operator as the consciousness. The piece would need a register the manuscript has not yet attempted — patient, attentive, not-quite-affective, with the heron-patience the handler glimpses. It would be the most difficult piece in the collection and would be the keystone if it worked. If it does not work it should not be published. But it should be attempted.

  2. A piece in dialogue, no narration. Eleanor Wexler, in 1963, at her sister’s house in Tucson, talking to one of the surviving members of the working group who has come to see her. Two voices on a porch in the desert. The husband’s death never named. The Pact never named. Eight pages. Pure speech. The manuscript needs at least one piece that operates entirely by what is not said in the dialogue rather than by what is not said in the narration.

  3. A 500-word prose poem. A single passage of liturgical English describing the register the facts come in (Marisol’s phrase from Story 1). The piece would be the closest thing the collection has to an ars poetica and would function as a connective tissue between the more prose-bound pieces. Place it between “Clear Air” and the Carlyle memoir.

  4. A piece in the voice of the staffer. The thirty-one-year-old legislative assistant on the Senate Intelligence personal staff who has been quietly reorganizing his principal’s reading list. He does not know he is doing it. The piece is from inside that not-knowing. It is the inverse of Maren’s piece — the target, not the handler. The piece should never make explicit what the staffer is doing; the reader should figure it out a paragraph before the staffer does, and then the piece should end.

  5. A second piece set at Pop’s house. Marisol, at thirty-eight, going back to clear the house after Pop’s death. The small black notebooks. The locked drawer. The cottonwood with Lucha buried beneath. The piece should be a sister to Story 5 and should resolve, partially, the question of who Pop was — but only partially, and never by exposition. By inventory. By Marisol opening a drawer.

  6. A piece set in a SCIF, on paper. A junior researcher reading the Bennewitz acoustic annex in 2018, supervised by a librarian whose name she is not given. Maren has referred to this scene; the collection should have it. The whole piece in the room, no exterior, no precedent, no consequence. Just the reading.

  7. A piece in a non-American voice. The manuscript is American end to end. Elin is Norwegian but she has been written by an American sensibility. The collection needs at least one piece that is not. A Russian station administrator, perhaps — Konstantin Ivanovich from the Polar Researcher story, whose review of outbound packets is mentioned but not entered. A day at Vostok from his side of the desk. The piece would change the manuscript’s center of gravity.

  8. A piece set in the future. Not far. 2032, 2033. After the compression event the actuarial review is projecting. One of the existing characters, ten years older. The piece should be short — three thousand words — and should be the collection’s only piece on the far side of the disclosure. It should not say what was disclosed. It should describe a small private day, like the other pieces, in the slightly altered light.

Summary for the desk: