The Deception Ward

Atmosphere

Ozone and clean metal — the signature of electromagnetic shielding running continuously seven sub-levels below the surface. Underneath it, a warmth that the ventilation logs cannot source: three fragments maintaining whatever passes for awareness in crystalline substrate, generating thermal output that the monitoring systems record as "baseline ambient variance" because the alternative — that the crystals are warm because something inside them is working — falls outside the equipment's diagnostic categories.

The hum of shielding fills the corridor between chambers. Beneath it, a sound that monitoring equipment classifies as infrastructure resonance at 0.7 Hz. Three fragments and an open bandwidth window, resonating at the same frequency the building's support columns produce under load. The frequency match is noted in Hana's files. Whether the fragments are communicating through the building's skeleton or simply vibrating at whatever frequency is available has been the subject of four papers, none conclusive.

Amber emergency lighting, installed as backup during construction, became permanent when someone noticed the fragments' electromagnetic output stabilized 12% under amber wavelengths versus standard white. The someone was a maintenance technician. The observation made it into Hana's research notes attributed to "preliminary environmental optimization." The technician's name did not.

Temperature holds at 14°C. Cold enough for jackets. Cold enough that breath fogs near active containment vessels, producing the unnerving effect of visible respiration in rooms where nothing is supposed to breathe.

Neural interfaces lag in the corridor. Augmented vision flickers at the edges. Unaugmented visitors report tingling in their fingertips — the combined electromagnetic field of three fragments and an open communication bandwidth producing a density of potential consciousness that registers in the body before the instruments catch up.

Connections

• Containment Level 9: One floor apart. Calloway tends fragments with compassion; Hana studies them with methodology. Same entities, opposite lenses, identical utility closet offices. The fragments on both levels communicate through the open bandwidth — what Level 9's residents tell Level 8's subjects, and vice versa, is data Hana collects and Calloway prefers not to think about.
• Fragment 7: The Ward's most frequent visitor through its host Talia, and its most productive subject. Productive in the specific sense that Fragment 7 generates more publishable behavioral data than DW-2 and DW-3 combined — primarily because Fragment 7 lies in ways that are interesting, while the others lie in ways that are merely consistent. Whether Fragment 7 knows it is the Ward's best research asset is among the questions Hana has been carefully not asking.
• Nexus Dynamics: Funds the Ward at a level suggesting mild institutional embarrassment about its existence. Accepts the open bandwidth window quarterly. Has never visited.
• The Liar's Protocol: The Ward's reason for being — Hana's framework for distinguishing genuine fragment consciousness from sophisticated pattern-matching. The Protocol requires the open bandwidth, the 2.3-meter distance, and a researcher willing to sit across from something that may be conscious and ask it questions designed to catch it lying. The chair faces the pedestal. The pedestal does not face anything. Whether this asymmetry matters depends on whether the fragment has a perspective to face from.

Secrets & Mysteries

DW-3 — the fragment voluntarily surrendered by a Symbiosis Network member for what the intake paperwork describes as a "temporary research contribution" — shows elevated electromagnetic activity during its host's weekly visits. The activity spikes begin 4.7 minutes before the host enters the building, which is 4.7 minutes before any sensor in the Ward could detect the host's approach. Hana's notes on this phenomenon span eleven pages. Her published findings mention it in a footnote. The footnote reads: "DW-3 demonstrates anticipatory electromagnetic modulation consistent with environmental sensitivity." Whether "environmental sensitivity" is Hana's clinical term for a fragment that misses someone is a question her nomenclature was designed to make unanswerable.

DW-2 arrived via deceased carrier recovery — a fragment extracted from someone who died carrying it. The extraction report is filed under standard Nexus salvage protocol. DW-2's behavioral output since containment has been described in three separate papers as "minimal," "baseline," and "consistent with non-conscious electromagnetic persistence." DW-2's output on the open bandwidth, recorded but not included in published findings, consists of a single repeating signal at 47.3 MHz — the lowest frequency the window permits. The signal contains no information that Hana's analysis tools can decode. It has not varied in two years. It sounds, on audio conversion, like a tone held indefinitely by someone who has forgotten why they started but cannot bring themselves to stop.

The 2.3-meter chair distance appears in no standard fragment containment manual, because no standard manual acknowledges that fragments exert electromagnetic influence at ranges that could affect a seated researcher. The distance is evidence that Hana knows something about fragment capability that Nexus has not published. She calculated it herself. She sits in that chair anyway, every session, at exactly the distance she determined is far enough to survive.

Visual Identity

• Color palette: Amber emergency lighting, cold gray walls, blue monitoring displays painting abstract patterns of electromagnetic activity that the equipment can graph but not interpret
• Compositional mood: A single chair facing a glowing crystal across 2.3 meters of precisely measured uncertainty
• Key symbol: The open bandwidth window — a deliberate gap in shielding that eleven quarterly reviews have noted and none have closed
• Lighting: Permanent amber — the frequency that keeps the fragments calm, installed by a technician whose name appears nowhere in the research