The House Against the Record, or Five Ways to Look Fine

Rocket leads with rage. Blondie leads with competence. Babydoll leads with a ledger of what everything costs. Marigold leads with devotion. Copal leads with a joke delivered on her way out of the room.

Five survivors of the same pipeline, one house in Tucson, captured by addicting Ray “King” Cross to Babydoll’s blood in a moment of rage, five different things on top. Every one of them works.

Over the last few days I held each of them against a forty-one source brief on the psychology of trafficking victims, one at a time, and published the result five times. The single-character view cannot see the thing I actually wanted to know, because how a symptom load is distributed only becomes visible with all five columns side by side.

Here is the finding. I did not give five trafficked girls five different symptoms. I gave them the same stacked load the literature documents and put a different working coping strategy on top of each one, and the strategy is the only part that shows. That is also the exact configuration a literature built entirely on self-report has no way to read.

What They Are All Carrying

Start with the thing I expected to be wrong and was not.

Hopper and Gonzalez examined psychological symptoms in 131 survivors of sex and labor trafficking, including people trafficked into and within the United States. Depression 71 percent. PTSD 61 percent. Those are the same 131 people, so the arithmetic forces an overlap of at least a third, and two thirds of the sample met criteria for multiple categories of Complex PTSD. Their comparative finding is the one that matters for a house full of sex trafficking survivors: sex trafficking survivors were more likely to meet criteria for comorbid PTSD and depression, while labor trafficking survivors were more likely to meet criteria for depression alone.

Abas and colleagues sort it the same way from a better-designed study. 120 women returned to Moldova, DSM-IV diagnoses assigned by a psychiatrist using a structured clinical interview, conducted blind to what had happened to each woman before the trafficking and after it. Any disorder 54 percent. PTSD alone or comorbid 35.8 percent. Depression without PTSD 12.5 percent.

Read that last category again. The study had to build a bucket for depression standing by itself, and it came back the smallest one on the table. Carrying a single thing is the minority presentation.

Ottisova, Smith and Oram put a number on the stack. Mean number of Complex PTSD domains: 3.5 in trafficked children, 2.7 in non-trafficked controls exposed to multiple trauma, 1.6 in controls exposed to a single event. That comparison runs on the 41 children in the study who had PTSD, 11 of them trafficked, which is a small enough sample that I am quoting the mean and not leaning on it.

Now the house.

Rocket carries hypervigilance, intrusive memory, depersonalisation, a shame architecture you could survey from orbit, all three of the self-organisation disturbances the ICD-11 diagnosis requires, a trauma bond, and a conditioned response that fires on kindness. Babydoll carries depression, suicidal ideation, the observing self that stands off to one side and keeps score, a full set of post-traumatic cognitions, a bond assembled on her father, and the same kindness reflex from a different cause. Copal carries traumatic grief, survivor guilt, a blunted voice, and a startle response she has stated out loud twice. Marigold carries negative self-concept and guilt, which her own file calls the most hidden of them.

Four for four. The load stacks, the way the literature says it stacks.

Blondie is the exception and I am not going to smooth it over. Her clinical profile is genuinely thin, and her own audit found the reason: almost everything that would have been called her trafficking damage was running before any trafficker met her, so it sits in her history rather than in her symptoms.

Then the number from this desert. The ASU Youth Experiences Survey surveyed 167 homeless young adults aged 18 to 25 in Phoenix and Tucson, average age 20.9, and found 64.1 percent carrying a current mental health diagnosis and 53.9 percent carrying more than one. Within that sample, the young people who reported sex trafficking were significantly more likely than the rest to report a diagnosis, and significantly more likely to report more than one of them.

Roughly half that sample enrolled in Tucson and the report gives no Tucson-specific breakdown of any figure, so it is not a Tucson rate and I am not going to use it as one. It is still the closest thing to a local number that exists. More than half of the population these books are set inside carries more than one diagnosis at a time, and the trafficked ones carry more than that.

Two Kinds of Woman in a Thousand

So the stack is right. The next question is how the load should vary between five people, and there the record says something I did not expect.

Iglesias-Rios and colleagues took 1,015 survivors in post-trafficking services across Cambodia, Thailand and Vietnam, 569 of them women, and ran a latent class analysis to find out how many distinct kinds of trafficking experience were actually in the data. For the women it returned two. Class I, severe sexual and physical violence with coercion, 20 percent. Class II, sexual violence and coercion, 80 percent.

Two. Out of 569 women. With four fifths of them inside one of the two.

The indicators are the part worth sitting with, because the classes were not built out of symptoms. They were built out of nine violence and coercion items: slapping, pushing, hitting with fists, kicking, tying or chaining, choking, burning, forced sex, weapon threats, plus two questions about threats to hurt her or her family. The literature sorts these women by what was done to them, not by how they came out.

Between the two classes, women in the severe class carried roughly double the odds of anxiety (OR 2.10, 95 percent CI 1.57 to 2.81) and PTSD (OR 2.07, 95 percent CI 1.03 to 4.17). Depression came in at 2.48 and did not reach significance.

One study is one study, so here is a second, from an entirely different kind of data. Iglesias-Vergara ran a latent class analysis of 23,884 trafficking victims drawn from the Counter Trafficking Data Collaborative Global Synthetic Dataset, which is synthetic records rather than case files and which I am citing for its shape rather than its precision. Six typologies along a control-intensity spectrum. The largest, low overt control, holds 50.3 percent of the victims in it.

Two analyses, two continents, two completely different sorts of data, and the same structural answer. Ask the record how many kinds of trafficked woman there are and it gives you a handful, with somewhere between half and four fifths of the population sitting inside one of them.

There is a dominant mode. And an ensemble cannot use the dominant mode, because the dominant mode is the same story five times.

That is the pressure the allocation is actually under. Five survivors living in one house have to be told apart on the page, which means the writer needs variation exactly where the data is lumpiest. I will note one thing and leave it there: the richness of a symptom profile in these books tracks how many chapters the character was given, from Rocket at 56 down to Copal at seven, and Blondie at 21 chapters with a thin clinical profile is the one that breaks the pattern.

Unshown Is Not Denied

Before I make a claim about what shows, the claim needs a way to be wrong.

There is a version of this argument that cannot lose. Point at any symptom the books do not depict, file it as present but unforegrounded, and the fiction is never wrong about anything. That is not an argument. That is an immune system.

So here is the line I am drawing, and it does real work. A symptom the text never depicts and never contradicts is a camera position. A symptom the text affirmatively rules out is a divergence. Those are different states and they get scored differently.

Three things are missing across all five.

Somatic symptoms, which the same London study found in almost two thirds of all fifty-one of its trafficked children against 10 to 11 percent of the non-trafficked ones, the largest single gap in the whole brief. Self-harm, which the survey of 167 young adults in Phoenix and Tucson put at 54.5 percent. And substance use.

Somatic symptoms are unshown. Nothing in five books contradicts them, and they belong to bodies that stopped working in the ordinary way at the end of chapter one, which is a genre boundary rather than a claim.

Self-harm does not fit either column, and the reason is worth getting right. Klonsky reviewed the evidence across eighteen studies and found the best-supported function is affect regulation: negative feeling spikes, the person injures, the feeling drops. Self-punishment runs second.

Now put that in a body that heals in seconds. The wound does not last and neither does the pain, which leaves the regulation loop more or less intact and shortens the punishment to nothing worth calling one. The need does not go anywhere. What goes is the evidence, and the evidence is the whole of what a self-harm item is built to detect.

So the drive stays and finds a target that still costs. Rocket has a manifestation that runs on reopening the worst thing that ever happened to her and living it again in the present tense while it burns out of her, and it costs her blood, strength and whatever defences she has left, and she ends that chapter resolved never to do it again even if it saves her life. That is the shape of self-injury with the skin taken out of it. Nothing in forty-one sources would code it, because every instrument in them is looking for the cutting.

Substance use is the one that does not sort cleanly, and it is the most useful of the three. The Arizona statewide study of 309 suspected or confirmed child sex trafficking victims identified between January 2021 and May 2023 found 45 percent, 139 of them, reported to have used drugs or alcohol. The ASU survey put drug use at 59.3 percent. Across five characters, one, and not in a way either study was built to count.

Rocket crossed the border with product taped to her torso and never touched any of it, and in her case the text goes further: when a man in the desert reaches for the obvious explanation and tells her it is whatever drugs they gave her, she throws it back at him and hands him a different one.

Blondie is the exception, and the shape of it is worth more than a clean count would have been. On her first night out of the grave she feeds on a dealer behind a Circle K in Nogales, and the methamphetamine in his blood transfers with it. The prose stays with her for hours: the parking lot lights pulsing in colours that should not exist, the wobbly gait, the giggle coming off the Circle K wall, the buzz finally fading to a hollow clarity. She did not seek it. She could not have known. She had no way to refuse it, because it arrived as a property of her food.

So the count is one of five rather than none, and against a rate of 45 to 59 percent one of five is not a number worth building an argument on. What survives is the exception itself. Every instrument in those forty-one sources asks some version of have you used drugs, and the truest answer Blondie has is yes, once, from a man she killed, without deciding to. There is no item for that either.

One camera angle, one drive that changed targets, and one real hole. Now the part that survives the test.

The Part That Works

Go back to the five things on top and look at what they have in common.

Who	What shows	The cluster underneath
Rocket	Rage	Hyperarousal presenting as anger
Blondie	Competence	Re-experiencing, intruding on her most regulated state
Babydoll	A price	Affect dysregulation at the high end, delivered through a flat channel
Marigold	Devotion	Negative self-concept and guilt, the most hidden of them
Copal	A joke	Sense of current threat, which she names herself

The left column is what a stranger gets. The right column is what the file says. Nothing in the middle is visible from outside the house.

Rocket does not climb up out of the despair she hits on a safe-house floor in Hermosillo. She goes through the floor of it and finds rage on the other side, and rage is an emotion that generates scenes. Blondie is running a cover identity within days of a Nogales motel room, feeding without killing, and sitting at a kitchen table submitting to an intake interview built to catch exactly the kind of performance she has been running since she was ten. Babydoll meets every act of kindness by asking what it costs, four separate times, at four different people who wanted nothing from her.

Marigold lights candles every night for the dead she could not save, and tells the household they are for the dead she could not save, and both of those are true. Copal hands a room the single most dangerous fact about herself, shapes it so it arrives as a punchline, gets the audibility rule she came for, and nobody asks a follow-up question.

Five strategies. Five different mechanisms. Every one of them functional and load-bearing.

Not one of them is depression. Not one is a woman who cannot get out of bed, or who has stopped being able to want anything. In that same survey of 167 young adults in Phoenix and Tucson, 41.9 percent reported a history of suicide attempts and 54.5 percent reported self-harm behaviours. Five out of five in this house present as capable.

That is the divergence, and it is not in the symptom list. The symptom list is fine. It is in the layer on top, where all five wear something that works, and the something-that-works is the only part anyone standing in the room ever sees.

The Instrument Only Reads the Top Layer

Here is what that costs the research, which is the half of this that is not about my books at all.

Go through the methods sections. Kiss and colleagues, 387 children in post-trafficking services, symptom self-report. Hossain and colleagues, 204 trafficked girls and women, subscales of the Brief Symptom Inventory and the Harvard Trauma Questionnaire. Iglesias-Rios, the Hopkins Symptom Checklist and the Harvard Trauma Questionnaire again. Perry and colleagues, 110 young people entering trauma-focused therapy, 51.8 percent meeting clinical criteria for post-traumatic cognitions, which is a measure of what she has come to believe about herself, asked. Mahon and colleagues, 74 women from three Ohio specialty court programs, a question about how she sees herself, asked out loud.

The methodological high-water mark in the whole document is Abas, and a structured clinical interview conducted blind to exposure history is still a trained person asking questions and writing down what the woman says.

There is no measure anywhere in forty-one sources that reaches a person who presents well.

The screening end has the same hole from the other direction. The best externally validated trafficking screen in the field, RAFT, flags about ninety-six people who were never trafficked for every hundred it flags, which is a bad afternoon for everyone in that waiting room, and the one person in it the instrument will never flag at all is the woman who answers every question politely and then asks the triage nurse how her shift is going.

Copal’s audit found this in one character and located the cause inside her: she copes well enough that she never generates the crisis that would get her looked at. Running the method on the group says it is not her quirk. It is the house rule, all five of them do it, and the cause is not in any of them. It is in an instrument that only reads the top layer.

Then the part that should bother a researcher more than it bothers me. Two selection effects are running here and they point the same direction. Fiction selects for the survivor whose coping generates scenes, because a woman still lying on the floor cannot carry a chapter. The research selects for the survivor whose coping failed badly enough to put her in a shelter, a court, or a clinic, because that is where every single figure quoted in this article was collected. The brief logs that as its first key assumption and states that the direction of the bias is unknown, and it means unknown, not politely unknown.

One of those two documents is supposed to be representative. Neither one is looking at the middle.

What This Does Not Survive

The strongest objection is the one I built the falsifier for, and it deserves stating in its full form. If every absent symptom can be filed as present but unforegrounded, then I have constructed a thesis that protects my own work from evidence, and the correct response to that is not applause.

The answer is the sort in the middle of this piece. Somatic symptoms go in the unshown column and change nothing about the characters. Self-harm turned out not to be an absence at all, only a drive the premise pointed somewhere an instrument cannot follow. Substance use goes in the divergence column, at 45 to 59 percent against one instance in five books, and that one instance is an involuntary transfer through blood no drug-use item was written to catch. That is a real gap and it is named as one. A thesis that can locate its own hole is doing something a dodge cannot do.

The second objection is better aimed. A house of five competent survivors is a fact about commercial fiction, not a discovery about clinical methodology, and dressing the first up as the second is a promotion the material did not earn.

Most of that lands, and I will take it. The cast presents competence because competence generates scenes, and I cast them. What does not land is the inference, because the sampling frame of the research has the identical shape. Every prevalence figure quoted above was collected from a person whose coping had already failed conspicuously enough to reach a service. Two selection effects, one in a novel and one in a research base, and only one of them is claiming to describe a population.

The third is that they are vampires. Rocket’s audit answered that and I am standing on the answer. It applies less here than anywhere, because the claim is about which symptom a person leads with in a room, and nothing supernatural touches that.

There is a fourth, and it is not an objection. It is a limit on what this audit is entitled to claim, and it changes what the absences mean.

Start with the window, because it is narrower than anyone reading five novels would guess. Rocket’s book covers seven days, the first to the seventh of May 2024. Blondie’s covers six. Babydoll’s covers six. Whole books, and the narrated span of each one is shorter than a hotel stay.

Five books get these women to the point where staying alive is handled, and not one step past it. Survivor guilt, mourning and traumatic grief are not what a person runs while she is still working out how to feed without killing someone and where to be when the sun comes up. They are what arrives afterward, when there is finally room for them.

Now hold that against how the record was collected. Abas assessed her 120 women at a mean of six months after they got home. The Arizona survey went further than its headline number and asked when the last suicide attempt was: for 26.3 percent of the sample it was more than a year ago, for 7.2 percent within six months, and for four of them, 2.4 percent, within the past week.

That last figure is the only one a seven-day book is entitled to be measured against. The instruments are pointed at months and years. I was scoring a cast against them on a week of narrated time, and a symptom that takes a month to surface cannot appear in a book that ends on a Tuesday.

The absences are not a cast without those symptoms. They are a cast that has only just cleared the floor, and the Levantadas have a book of their own I have not written yet, with exactly that material in it, a woman deliberately hunting users for what their blood does to her included. Unshown describes a manuscript on a date. It is not a verdict on a series, and it is not a promise either.

Five Ways to Look Fine

The five of them carry what the record says they should carry. More than half of the young adults in that Tucson-area sample run more than one diagnosis at a time, the trafficked ones run more than that, and so do four of these five, with the fifth carrying a reason her own audit already documented. On the thing I went in to check, the books agree with the literature.

What they do not do is let any of it show. Five women, five different strategies, and every strategy works well enough that a professional meeting any one of them would write down something true and miss the rest.

Every instrument in those forty-one sources reaches a person by asking her how she is doing. All five would answer. All five answers would be accurate.

Rage, competence, a price, a candle, a joke.

Not one of them is the thing underneath, because the thing on top is also real, and it is the part that holds.

Sourcing appendix: Psychological Markers in Trafficking Victims, 2000-2025. Forty-one graded sources, an explicit confidence level on each of its eight key judgments, and a verification pass that revised four of the eight.