The form has forty-two fields. Fourteen of them are for dates.
Reyna has used this form for nine years. She knows which fields get left blank (the optional ones, always), which get filled wrong (Field 7, almost always), and which get partially filled because the person ran out of language mid-sentence and never came back to finish. Field 7 is the one she's thinking about now, at eleven in the morning, the Recalled intake office mostly quiet, a reconsolidation crew from the Wayne State clinic parking their van in the lot below her window.
Field 7: Date of departure from previous residence.
It asks when the body left. The body is the easy part.
Lena had said: I don't know where I'm from in the same way since my mother died.
She said it on Tuesday, in the one-on-one Reyna approved for week four — individual format to go deeper, not to go around. Reyna had written it verbatim into her session record immediately, because verbatim is not interpretation. Verbatim is the thing itself.
Then she had added, below the verbatim, in brackets: [Note: Verify whether Lena has undergone reconsolidation for grief-related memory. If yes, the timeline below cannot be taken as organic.]
This is the note that is making her look at Field 7 again.
Lena was from Detroit and from Akron. Detroit was her mother's city. Akron was the city she moved to at twenty-three. Her mother died in Detroit, eighteen months before Lena left Akron. In those eighteen months, Lena had lived in an Akron she did not know how to call home anymore.
Lena had also, six months before her mother's death, undergone reconsolidation treatment for workplace-related trauma. Insurance-mandated, first-line. The standard intake: reactivation cue, labile window, Gen-3 array, extraction, erasure of the acute distress valence. Lena still had the memory of the incident. She just didn't have the same charge attached to it.
The question Reyna cannot answer, looking at Field 7, is this: did the treatment contaminate the timeline?
The labile window doesn't always stay labile. Sometimes adjacent memory traces destabilize alongside the target. Sometimes what comes back from reconsolidation is cleaner than what went in — stripped of noise that included meaning. Lena's workplace memory had been treated six months before her mother's death. Which meant that six months before her mother's death, Lena's attachment geography might already have been shifting. The city that felt less like home after the treatment might have absorbed less of the subsequent grief. Or more. Or differently.
The form asks Field 7: When did you leave? It means: when did the body cross the state line?
It doesn't know that it is also asking a neural question.
Reyna opens a new document. Types:
PROPOSED FIELD 7b: Approximate date when the previous residence stopped functioning as home (may differ from physical departure date).
Looks at it. Soft. People don't trust forms that ask how they feel. She revises. Then adds:
PROPOSED FIELD 7c: Has the client undergone reconsolidation therapy affecting memories of the previous residence or primary attachments? If yes, provide approximate date and scope.
Field 7c is the field that didn't exist before 2034. Before 2034, it wasn't necessary. Before 2034, you could assume that when someone said I stopped feeling at home eighteen months before I left, the timeline was biological — grief, or job loss, or a neighborhood that changed — not a technician's choice about which memories to stabilize and which to let run.
Now you can't assume.
This is what Reyna means when she says the form was built for a different city.
Detroit has eleven reconsolidation clinics. The three-block corridor around Wayne State runs on experiential data revenue the way it used to run on automotive parts contracts. The van from the Wayne State clinic is still in the parking lot below. Through her window she can see the tech pulling the equipment case out of the back — the Gen-3 array portable unit, padded case, the wireless rig that uploads to the exchange in real time. The van comes to the client, not the other way around. The clinic is mobile because the protocol requires the client to be relaxed, in a familiar environment, not anxious from transit. The infrastructure of extraction has been organized around the labile window's sensitivity to stress. It travels. This morning's extraction will fuel AI systems she will never interact with directly. She understands this abstractly, the way you understand the grid: the lights come on when you flip the switch, and somewhere upstream someone is paying for the electricity.
She knew a woman who worked the extraction shifts for three years at the clinic on Woodward. What she learned from that conversation: the clients don't look different after. The session takes forty minutes. You put on the array, you're given a reactivation cue, the labile window opens, the tech monitors the oscillation patterns on the readout, the extraction happens in the background while the therapeutic protocol runs in the foreground. Most clients describe feeling clearer afterward. Some describe feeling slightly less themselves, in a way they struggle to articulate. The form they sign before the session has forty pages. The form does not ask: Do you understand that the tech will upload your distress to a server in Dallas?
The form also does not ask: If we edit this memory, and you later lose the anchor it was attached to — a mother, a city, a sense of where you belong — will the loss land differently because we changed the architecture first?
The third referral this month mentioned her mother during intake. Reyna had noted it — third, mother, mentioned — without connecting them. One is an instance. Two is coincidence. Three is something she needs to sit with before naming.
All three had crossed state lines after a primary loss. All three had undergone reconsolidation, verified by the insurance trace in their records. All three had filled out Field 7 with the date the truck was loaded, while describing an internal departure that didn't match.
All three, when asked where they were from, had paused.
She has worked in intake long enough to know that there are different kinds of pauses. The pause when the answer is complicated. The pause when the answer has changed recently. The pause when the person is trying to decide what version of the answer to give. And a fourth kind, which she has only started noticing in the last two or three years: the pause that looks like searching for a fact that used to be automatic. The pause that means: I know this. I used to know this without thinking. I'm not sure why I have to think about it now.
She writes it in the parallel log — not the intake database, not the official record, just the track she keeps when she's noticing something she can't prove yet:
Three cases. Primary loss following reconsolidation treatment. Intake pause on geographic identity question. Not explained by grief alone. Need twenty more before I say anything to anyone.
The contamination literature says the labile window doesn't always close cleanly. Adjacent traces can bleed. The Okafor decision requires disclosure, but disclosure is not the same as understanding. You sign the form. The form tells you adjacent traces may be affected. You sign it because the insurance won't cover the cognitive therapy otherwise, and the cognitive therapy is what you need for the thing that happened at work, and you are not thinking about your mother, because your mother is alive and in Detroit and has nothing to do with this.
Six months later your mother dies.
Now you are thinking about your mother. And you are thinking about Akron, and whether Akron is home, and whether the feeling that Akron stopped being home has something to do with your mother dying or something to do with what the tech did to the architecture when you were focused on something else. And the form asks you Field 7: When did you leave? And you write June because that's when the body left, because that's the only thing you're certain of.
That is Lena. Or a version of Lena that Reyna is constructing from incomplete data, which is always the risk. This is why she writes it in the parallel log and not in the intake record. The intake record is for facts. The parallel log is for hypotheses she's not ready to abandon.
She goes back to the proposal document. Field 7b. Field 7c. She adds a rationale section:
In the current instrument, physical departure date (Field 7) is the primary temporal marker for residential history. This field was designed for a population whose memory timelines are organic — where the subjective experience of leaving tracks closely to the physical event of relocation. In the current clinical environment, a significant proportion of clients have undergone reconsolidation therapy that may affect attachment memory architecture. In such cases, the internal departure — the subjective experience of ceasing to feel at home — may precede the physical departure by months or years, and may not be fully available to introspection in its original form.
Proposed fields 7b and 7c would allow intake workers to document this temporal gap and flag cases where reconsolidation may have affected the timeline. This is not diagnostic. It is documentary. The form cannot reconstruct what the therapy did. It can note that the therapy happened and hold space for the gap between dates.
She reads it again. Accurate. Careful. She is already writing in the language of committee survival rather than the language of the parallel log, and she notices this, and decides it is the correct choice. Will probably take a year to get through the design committee and three more years to pilot, assuming approval, assuming the political environment around reconsolidation doesn't shift so far that the whole conversation becomes impossible.
She saves it. Draft 1. She has no illusions about Draft 1.
But she made it exist. It did not exist before this morning. Now it does.
Lena's verbatim statement is still in her session record.
I don't know where I'm from in the same way since my mother died.
The operative phrase — Reyna keeps returning to it — is in the same way. She still knows where she's from. She just can't access it without effort. She lost the automatic version. What's left is the effortful version, which is technically the same information and experientially something else entirely.
Reyna does not know if the therapy changed this. She does not know if it accelerated a grief process that would have happened anyway, or opened something that would otherwise have stayed closed, or did nothing to the attachment geography and the loss of the automatic is just what losing a mother does, full stop, with or without Gen-3 arrays and extraction fees and forty-page consent forms.
She does not know. That is the correct answer.
What she can't accept is a form that pretends it doesn't matter.
The van from Wayne State is pulling out of the lot. The tech drives the extraction unit somewhere — the exchange, or the next appointment. Somewhere downstream, this morning's session is being labeled, segmented, verified-pure or contaminated-disclosed, priced and sent to run in systems that reason by borrowing what people feel.
The people who feel it don't know their geography is fuel.
Field 7b. Field 7c. Draft 1.
She goes to get coffee. The form has forty-two fields. It will have forty-four.
She thinks about the parallel log, walking to the break room. The log is personal, unofficial, held in a system that doesn't talk to the intake database. She started it two years ago when she realized she was noticing things that didn't have a field. Not anomalies. Not errors. Just — texture that the record couldn't hold. The feeling of a pause. The particular way someone qualified their answer. The fact that three people in one month mentioned their mothers.
None of this would stand in a case review. She knows that. The parallel log is not evidence. It's the place she puts the questions that precede evidence — the shape of something she doesn't yet know how to see directly.
She pours the coffee. There is a readout on the break room wall, the clinic's intake volume for the month — a dashboard the Wayne State system pushes to partner facilities as part of the data-sharing agreement. Forty-seven sessions in March so far. Thirty-nine with extraction consent. Thirty-one verified-pure. Eight disclosed-contaminated.
She reads it the same way every morning. The numbers. The ratio. She has never asked herself, until this moment, what happens to the clients in the disclosed-contaminated column when they eventually show up at an intake desk like hers. What they say when Field 7 asks them when they left.
What their pause sounds like.
She takes the coffee back to her desk. Opens the proposal document again. Beneath the rationale section, she adds a final line:
Note: This proposal does not address the question of whether contaminated memory traces affect geographic attachment differently than uncontaminated traces. That question is outside the scope of this instrument. It is not outside the scope of the problem.
She saves it. Draft 1, with the line that marks the edge of what she can say yet. The line that says: I see the shape of something. I am not naming it until I have twenty more cases. But I am putting it on the record that I see it.
The form has forty-two fields. The parallel log has no fields at all.
Both are necessary. That is the thing she knows.