This chapter argues that the ten Home records each begin with one occupant's feeling and that each record's doubt about itself brings in what the feeling left out, in three ways that follow where the home is kept. What it adds to the argument between the home as a refuge and the home that others share, build and pay for concerns framing: a catalog that must find fault with every question finds the fault on the side of the building, of the other occupant, or of a hazard, depending on where the question puts its walls. The method comes first, briefly; the three sections after it read the records.
1.1Records and method
The records come from the public catalog of Cosmic Press, which the studio serves as a manifest and a set of numbered detail files. The catalog holds 768 entries under 24 topics, gives its source as "AI-generated examples", and every one of the 768 entries marks itself AI-generated. An entry has a question in the voice of a person, a compressed general form of it, a context, a need, an assumption, a line on what the compression keeps (preserved), a line on what it loses (lost), a counterquestion, a defect and six scale notes. It has no options and no judgment, and it never answers its question.
The studio files 31 entries under Home: 9 in the catalog's first 240 entries, 21 in the long middle stretch from q241 to q744, and 1 in the closing stretch. The selection rule is one sentence: every third Home entry in the catalog's order, counted from the second. It takes 10 entries, q044, q047, q050, q274, q346, q418, q490, q562, q634 and q706, of which 3 lie in the first stretch and 7 in the middle. The 8 earlier Cosmic Press studies read 97 entries between them, one of them a Home entry, q754; counted from the first, the same interval would take 11 entries including it. Counted from the second it takes none that an earlier study read, and the study's counting script stops with an error if any case it reads was read before. The fetch scrubs any field that names a model; it found none to scrub in the ten.
The stretches differ in manner. The three first-stretch records are short and flat: their questions run to 52 or 53 Korean characters, and none of their scale notes is a question. The seven middle-stretch records put all six scale notes as questions, 42 of the 60 notes in all; five of them run long, their questions reaching 88 characters, and one, q562, labels its notes Individual, Relationship, Society, Other lives, Planet and Distant future. Two long records set an English word in brackets inside the Korean, Home in q274 and Homing in q490. The two latest records, q634 and q706, return to small scenes and plain words.
Cosmic Press publishes its records in Korean. The English in this paper is a translation made once by the editorial session of this study, sentence by sentence, with nothing added; each case's English seed names the Korean original by its hash, and the appendix gives every field of the ten in that English. Every block quote is that English, word for word. The Korean seldom names the subject of a sentence; where English needed one, the translation used a passive, a gerund or a noun phrase where it could, and "I", "my" or "we" only where the Korean carried them.
Every count is made on the Korean by a plain word pattern named beside it in the study's numbers (numbers.md, made by scripts/count.mjs), so that the translation decides none of them. The question names another person in or by the home, a neighbour, parents, a roommate or a mother, in 5 of the ten; something sensed, wind, light, air or a wind chime, in 4. This paper assigns each record to one group by where its question keeps the home, and the groups are numbered in the numbers file. Two keep it against what lies past its walls and papers: q044, a neighbour's child heard through the building, and q047, an address the city cannot read. Four keep it among the people inside: q050, q274, q346 and q562. Four keep it in a sense: q418, q490, q634 and q706.
The counts then set the scene for each group. In 3 of the ten the record's criticism turns to the building, a structure or the dwelling's actual condition: q044, q047 and q562. In 4 the question put back turns to someone else in or near the home: q050, q274, q346 and q706. In 2 the criticism names a practical hazard, mould and allergies in q418 and too few streetlights in q490, and those two, with q274, are the 3 that charge sentiment, romance or subjectivity. In 2 the criticism finds the trouble in a word of the framing, and in 2 in what the compressed question or the setting can hold. Some field names a refuge, a nest, a shelter or a fortress in 4 of the ten, all from the middle stretch; across the topic the same words stand in 11 of the 31 Home entries and in 35 of the catalog's 768. The fourth scale note gives the question to an animal, a plant or a small life in 7 of the ten; two of the three first-stretch records speak of bodies there instead, and the third of a life in which help is given and received.
The script also takes the series' frame measure: for each field, the share of its five-word runs that recur in at least 5% of the records that fill it, and a field whose ratio reaches 30% counts as the form's words rather than the record's. Every field measures 0% in Korean, in the whole catalog and in the Home topic, and every field of the ten in English measures 0%. No field is set aside as frame. The form itself is fixed, however: every record must criticise its own compression and must put a question back. That a record finds fault with itself is therefore not a finding. What this paper reports is where the fault is found and on whose behalf.
All 10 contexts call their situation hypothetical, as do 759 of the 768 entries in the catalog. Every sentence quoted here is read as framing, not as evidence.
1.2Walls and papers: the building and the room
Two AI-generated records keep the home against what lies beyond it, a neighbour heard through a building and a city that reads a form. In both, the question asks about people, and the record's doubt moves the trouble to the material: a building that carries sound and a room whose condition no address can vouch for. The first is a hypothetical multi-unit building:
If sleep is broken by a neighbour's child's footsteps, yet the hope is that the child can walk about freely, how will we live together?
The question is already generous: it holds broken sleep and the child's freedom side by side and asks how to live together. The record sets it among hypothetical neighbours "in a multi-unit building where sound carries easily", and the assumption it finds in the question is a bargain between persons: "One side must keep quiet or the other side must endure." What the compressed form loses, the record says, includes "the state of the building's sound insulation". The building enters as a lost detail and stays as the main charge. The record's criticism of its own compressed question reads:
The word adjustment hands a structural defect over as the neighbours' responsibility.
The fault is found in a word. To ask how neighbours "adjust" their differing needs is to take a defect of the structure and give it to the people who live in it. The question the record puts back makes the same move as an offer: "Is it not possible to speak together of the building's problem rather than of habits of living?" In Cheshire, Easthope and ten Have (2019), what one household does to make its home can unmake the home next door; this paper reads the hypothetical record as giving that overflow a material channel and declining to settle it between the two households. Şentop Dümen and Rasmussen, Neighbour noise in multi-storey housing (2025), argue for disclosing the acoustic performance of buildings so that buyers and tenants can choose with knowledge. This paper reads the record's turn as running the same way, from conduct to construction, though the record proposes nothing; it only names the building as what the question forgot. One of its scale notes adds that the repeated loss of sleep "is not covered over with the universal ideal of living together". The record will not let the ideal of shared life, any more than the neighbours' manners, absorb the cost.
The second record moves from the floor to the form:
If there is a room but the address is hard to write down and has to be explained every time, am I properly living in this city?
Its compressed form asks "Who fills the gap between the dwelling as it is officially read and actual residence?", and its assumption is that residence "is recognised only by fitting into the set box for an address". Somerville, Homelessness and the Meaning of Home (1992), looks for the political meaning that official definitions, laws and state provision give to homelessness; this paper reads the hypothetical room as standing in the gap that article's title names, a roof without the recognition of a home. The record then doubts recognition itself as the measure:
Does the room become safe by the mere fact that an address has been made?
Reduction to a problem of recognition leaves out the actual condition of the dwelling.
The question asked to be counted as a resident. The record doubts that being counted settles the room's condition: an address does not by itself make a room safe. One scale note keeps the feeling apart from the fact, telling "the shame felt in explaining" apart from "the inconvenience in daily life", and another sees together "The recognition of residence and the function of a space that protects the body". In both records of this group, then, the doubt turns from what people say to each other, or what an official form reads, toward the building and the safety of a room. These two are also the records whose fourth scale note speaks of bodies, a young body moving and a body falling asleep, and a space that protects the body, where later records speak of animals. This section's contribution is that a turn from manners and recognition to materials, which this paper reads in Somerville and in the neighbour-noise argument, is already built into these two hypothetical questions.
1.3The people inside: the occupant the question does not speak for
Four AI-generated records keep the home among the people who share it: an adult child back with parents, roommates at odds over a living room, roommates at odds over a sink, and a mother whose children call her by her name. Here the record's doubt does not move to the building. It takes the part of the occupant whom the question's feeling does not speak for, and it does so even when that means turning against the record's own assumption. The first imagines an adult back in the parents' home who wants to set "my own time" and asks whether gratitude and independence can be together. The context gives the reason as "economic reasons", and the assumption is what the asker's wish runs up against: "If a dwelling is provided, the way of living must be followed too." A reader might expect the record to side with the asker's autonomy. Its question back looks the other way:
Does the freedom of my time not disturb the sleep and work of the people who live together?
Asking only about the boundary of control shrinks the mutual burdens of shared living.
In the study of multigenerational households by Easthope, Liu, Judd and Burnley (2015), what members felt as home was shaped by how much control each of them had, and each had a different amount. This paper reads the record's compressed form as asking only about control, the boundary "between support and control", and its criticism as saying that control is half the matter: the other half is the burden each member's freedom lays on the rest. Its scale notes keep the asker's side too, "A grateful heart does not mean having agreed to every demand", and end by setting "The time actually lived together" above "the abstract ideal of self-reliance". The record decides nothing between parent and child.
The share-house records show the same turn more sharply, because there the record's own assumption takes a side and its doubt then takes the other one. The first asks whether a roommate's hobby things and sprawled blankets make the house "a warm nest, or a common space encroached upon", and assumes an answer:
The belief that the essence of a home lies not in perfect observance of rules and in hygiene but in the traces and warmth of a comfortable life showing through.
The question back and the criticism then speak for the roommate that belief leaves out:
Does someone's sprawled comfort not come to another roommate as a burden of housework to be cleared away and as visual pollution?
The subjective emotional state called coziness can be misused as a pretext that justifies disorder in the occupation of space.
What the compressed form loses is both feelings at once, "The relief felt on lying down, as if collapsing, on the cluttered living-room sofa after work, and the irritation that welled up every time the things not tidied away were seen". The second share-house record turns the roles round. Here the question takes the side of order: a roommate promised to have the dishes in the sink "washed within 30 minutes" and left them neglected all night, the need the record names is "to enjoy a calm order inside the home at least", and its assumption favours that need:
The insight that the peace of a household stands precariously not on grand affection but on the honest keeping of extremely trivial promises of daily life, such as clearing the dishes from the sink and putting the shoes in order.
The doubt then defends the one who broke the promise:
Must a roommate who has come back after being worn down by all kinds of heavy work outside be pressed to the point of suffocation, with talk of the home's trust, for not washing a few dishes right away?
The risk of drifting into a cold disciplinarianism that, without regard for an individual's severe burnout and fatigue, enforces only a fastidious observance of rules.
Set side by side, the two records argue opposite cases and make the same move. In one the doubt protects the roommate who tidies against the warmth of the one who sprawls; in the other it protects the roommate who, worn out, collapsed into bed, against the order the question asks for. Heath's study of peer-shared households (2004) finds proximity, shared space, real and symbolic, and ritual important to the everyday framing of housemates' relationships. This paper reads the two hypothetical records as treating the shared space as the place where any one housemate's feeling of home becomes another's labour or another's strain, and as refusing, each time, to let the question's feeling be the whole of the home. Both also carry the home toward policy and machines in their notes: spatial buffers for public housing in the first, dishwashers as a required standard and a robot-run smart home in the second.
The fourth record is the hardest case, because its question is a celebration. At a hypothetical seventieth-birthday feast, the grown children call their mother by name instead of "Mother":
What is the tremor when a mother, who all her life was called only 'so-and-so's mum' and 'so-and-so's house' and whose own name was erased in household labour, is called at the table by the whole three syllables of her own name?
The record's doubt runs both ways at once. Its question back defends the role name the celebration sets aside:
Is being called 'Mum' tenderly within the family not a token of love worth more than any splendid title in the world?
Its criticism of the compressed form charges the celebration with doing too little:
There is a simplification that tries to resolve at a stroke the long violence of structural patriarchy, the erasing of women's names, with the emotion of a heartwarming family event.
Young's essay, House and Home (2005), holds both thoughts, that the comforts of home have come at women's expense and that rejecting the ideal of home wholesale is open to question. This paper reads the record as splitting that ambivalence between its two self-criticisms, one guarding the love in "Mum", the other refusing to let a single warm family event stand for the end of a structure. Brickell's call (2011) for a geography that helps address domestic injustice as well as showing it fits here, in this paper's reading, as the measure the event fails: it feels like addressing without doing it. Here the person the question does not speak for is the mother as a mother, and also the women whose names, in the record's words, a structure erases. The doubt reaches beyond the occupants to a structure, as the neighbour record reached beyond them to a building.
The group's contribution to the debate is a shift in where the critic stands. The critical works name who pays for the comforts of a home; these hypothetical records name a different payer each time, chosen by the question's own feeling: the parents' household, the tidy roommate, the tired roommate, the woman behind "Mum".
1.4Senses: the hazard, the sentiment and the words
Four AI-generated records keep the home in a sense: a damp floor dried and a wind chime ringing, a lit sign at the mouth of an alley, the air after watering, light on a table. In the two long ones, the record's doubt brings a hazard and a charge of sentiment against the pleasure the question describes. In the two short ones, the doubt mostly stays with the record's own words, and only once does it look to the rest of the house. The first is a hypothetical rainy-season afternoon in an old multi-household house. The need it names is to "make the house a dry, clear fortress of peace", and the assumption gives the home a breathing body:
The thought that a home is not a mechanical sealed box but an organic filter of the senses that takes in and governs the changes of the seasons outside and the wind, breathing together with them.
The record's doubt arrives from both directions. The question back offers the sealed box the assumption rejects: "Does shutting the windows tight and running a powerful dehumidifier and air conditioner 24 hours a day not give a far more scientific and perfect comfort?" The criticism names the cost of the romance:
The risk of worsening the real hygiene problems of indoor mould growth and of triggering allergies while enjoying ventilation and the romance of an analogue wind chime.
What the compressed form loses is equally plain, "The burden of the summer gas bill that has to be paid to run the boiler, and the small labour of wiping up the raindrops". Nobody else enters this home; what enters is damp, money and work. Dupuis and Thorns (1998) assess home as the setting of daily routines and a site of constancy; this paper reads the hypothetical afternoon as such a routine, a small ritual of control over weather, and the record's doubt as a reminder that the ritual has material consequences.
The second sense record moves the home's edge out into the street:
The insight that the place called home does not begin when the door of my room is opened but already begins when the familiar sign light at the mouth of the alley, the smell of the bakery and the feel of the familiar paving blocks are stepped upon.
Easthope, A place called home (2004), draws on a literature that treats home as a particularly significant type of place, giving insight into identity and into the conflicts and political economy of home-places; the hypothetical office worker's relief at the old dry cleaner's sign is home stretched to a street corner. This paper reads the record as naming a political economy in what its compression loses: "The anxiety of place loss to be gone through if that dry cleaner's closes or the sign is changed, and the threat of urban redevelopment." Then it doubts the feeling twice:
Is giving the grand meaning of the home's calm to merely one shop sign seen every day not excessive sentimentalism?
The concern that, resting content only in the familiarity around the home, the real risk factors of too few streetlights in the alley or of safety on foot late at night are overlooked.
The friendly light is set beside the missing streetlights. Atkinson, Ecology of Sound (2007), finds that the sound of a street can shut people out or quietly steer them; the record's street is lit and smelled rather than heard, so the lens fits only in part, but this paper reads the record as making a similar point about light: the sign that guides one walker home sits in an alley where the record warns of too few streetlights. Its notes carry the homing feeling far, to a sea turtle that senses the faint scent markers of the beach where it was born, and to an astronaut who plays a hologram of old streetlights aboard a ship voyaging for hundreds of years.
The two latest records are smaller and plainer, and their doubt changes. The first asks whether the air softens when the veranda pots are watered, "do only I feel this way?", and the record doubts the cause before anything else:
Might it not be because of that brief moment of opening the window and letting the wind in, rather than because of the plants?
In the compressed question, the term living space is stiff, so the lived feel of the veranda pots can fade.
The fault here is in the record's own vocabulary: "living space" is too stiff a phrase for a pot on a veranda. No neighbour, roommate or hazard enters its doubt. The last record keeps light on a table with the curtain "closed only halfway", and its criticism is again about what the framing can hold, "The position of the curtain alone can hardly hold the angle of the sunlight by season and the state of the eyes." Its question back, though, is the only one in the sense group that looks past the table to the rest of the house:
While the light is enjoyed, who should mind the darkness of the other rooms in the same house?
Mallett's review (2004) asks whether home is a place, a feeling or a practice; this paper reads the two late records as keeping home almost wholly a feeling, and their doubt as mostly a doubt about how well words hold a feeling, with one glance at the dark rooms of the same house. Across the group, the notes hand the home to a squirrel under a rock overhang, a sea turtle, the small roots in a pot and a sparrow at the window, and send it into a sealed dome and a spaceship.
This section's contribution is a limit on the pattern of the two before it. Where the home is a sense, the turn toward what is shared and material survives as hazard and cost in the long records, and thins in the short ones into a doubt about words and a single question about who else lives in the house. Across the chapter, every question began with one occupant's feeling and every record's doubt let something into the home that the feeling had left out: the building, the actual room, the other occupant, a structure, mould, a dark alley, a stiff word. That, read against the debate, is what these AI-generated framings add: the haven is posed, and the critic's correction is written into the same record.