January 2004 – October 2005 · Classified Facility, Scottish Highlands
The facility had no official name in any document above classification level six. The personnel who operated within it referred to it as the Nursery, which was the kind of institutional dark humour that develops in places where the work is serious enough to require some relief valve and grim enough to ensure the humour never gets very far from grim.
It had been built into the base of a mountain in the northwest Highlands during the early Cold War years, when the British government had a period of enthusiasm for facilities that would survive things that most things would not survive. The mountain above it was unremarkable — heather and scree and the specific Scottish quality of landscape that looks inhospitable because it is. The facility below was a series of interconnected concrete corridors and chambers at three levels of depth, maintained at a constant nine degrees Celsius by a ventilation system that also smelled, very faintly, of ozone and the particular chemical tang of a space that has been cleaned too many times to remember what it smelled like before.
There were ten candidates in January's intake. By February, four were gone — two medical withdrawals, two who had simply walked to the surface access and declined to return underground. By April, six remained. In August, a training accident that was classified and sealed removed two more. The remaining four continued.
Alen was one of the four. He was also, by the end of the first month, the reason the instructors had begun referring to the training assessments in two categories: Richard's results, and everyone else's.
The programme had no schedule in the conventional sense. It had a structure, which was different: a sequence of disciplines delivered in the order they were needed, at the intensity required by the specific candidate, without the fixed timetabling that would have made it legible to anyone observing from outside. What it looked like from inside was this:
06:00 hrs — Virology and Pathogen Identification
The room was twenty feet square, with a single screen at one end and a table at the other and nothing else. The lighting was calibrated to prevent eye fatigue over long sessions. The instructor was a small, precise man in his fifties whose name Alen was not given and did not ask for. He had, in a previous professional life, worked for Umbrella. He had left that employment under circumstances he described as a crisis of professional ethics and which the intelligence summary Alen had assembled on him suggested were somewhat more dramatic than that. He knew the viral taxonomy of Umbrella's primary research programs with the intimate familiarity of someone who had helped build it.
He placed images on the screen. He did not explain them. He waited.
The first image showed necrotic tissue — a forearm, the epidermis hardened and discolored in a pattern that Alen recognized from the theoretical literature and from four years of personal research and from something deeper and more molecular that he had never found a satisfactory term for. He ran the cellular signature against his internal index and the match came back before he had consciously completed the comparison.
"T-Virus. Strain Epsilon variant," Alen said. "The epidermal hardening is characteristic — the virus is using the keratin architecture as structural reinforcement rather than simply destroying it. It's not just infection. The host's own biology is being co-opted to create a secondary defense layer." He studied the tissue margins. "Standard small arms fire against this would be degraded by approximately thirty percent. The effective neutralization point is the spinal column at C3. Sever the motor pathway at that junction and the viral network loses its command architecture. The body stops functioning even if individual cells remain viable."
The instructor looked at him. It was the look of a man who has asked a question expecting a partial answer and received an answer that requires him to revise his question.
"How do you know the spinal junction?" he asked. "That is not in any published literature."
"It's the logical extension of the viral architecture," Alen said. "The T-Virus doesn't think. It optimizes. It follows the existing nervous system pathways because they're the most efficient route for systemic distribution. Which means the nervous system is also the most efficient point of disruption." He paused. "The specific junction is C3 because that's where the motor and autonomic pathways share the most structural density. One point. Maximum effect."
The instructor nodded slowly. He put the next image on the screen. He didn't comment on the previous answer. He had learned, by the third session, that commenting on Alen's answers produced less useful information than simply providing the next prompt and observing where the analysis went.
12:00 hrs — Asymmetric Close Quarters Combat
The CQC instructor was a former SAS staff sergeant named Brennan who had the physical build of something assembled primarily for impact and the movement economy of a man who has spent twenty years in environments where unnecessary motion costs you things you cannot afford to lose. He wore the full padded assessment suit, which he had needed perhaps twice in his career before arriving at the Nursery. He wore it every session now.
The drill was simple. Come at me. The objective was not to win — no candidate was expected to defeat Brennan — but to survive long enough and demonstrate sufficient technique to satisfy the assessment criteria.
Alen came at him.
What happened next was technically a fight in the same way that a scalpel operation is technically cutting. Alen moved under Brennan's opening haymaker with a body displacement so small and so perfectly timed that it looked less like evasion and more like the punch had simply decided to go somewhere else. He did not counter with strikes to the body mass — the obvious targets, the trained response. He went for the nerve clusters. A precise jab to the radial nerve junction at the lateral epicondyle. A follow-through kick not to the knee but to the femoral nerve corridor at the inner thigh, the point where the nerve bundle runs closest to the surface before disappearing under the muscle. Precise. Anatomically exact.
Brennan was on the floor in three seconds. His right arm was temporarily non-functional. He had not been damaged — the nerve strikes were calibrated to disrupt without injure, which was a level of control that required either years of dedicated practice or something else — but he could not use the arm.
Alen stood over him. He was breathing normally. He looked at his own hands for a moment with the particular expression he sometimes had — the expression of someone checking their own work against an internal standard they hadn't consciously formulated.
"Stop thinking about it as a fight," he said quietly. He was talking to himself, not to Brennan. Brennan had the distinct impression of overhearing something rather than being addressed. "A fight has an opponent. This is a system with failure points. Identify the failure points. Apply pressure. The system stops." A pause, barely audible. "Think like a virus."
Brennan, lying on the mat with a temporarily unusable arm, processed this and decided it was the most accurate description of effective combat he had ever heard, and also the most unsettling.
23:00 hrs — Psychological Resilience Assessment
The tank was a standard sensory deprivation unit — a sealed chamber, temperature-controlled water at body temperature, complete darkness, acoustic isolation from external sound. The programme used it for psychological stress assessment and, in its more advanced applications, as a tool for testing the relationship between sensory deprivation and cognitive clarity under artificial duress.
The artificial duress was the audio feed. Recordings from the Raccoon City incident — genuine recordings, sourced through channels that the programme did not document in its official methodology. Crowd noise. Emergency broadcast fragments. The specific acoustic signature of a city discovering what was happening to it. And underneath all of it, in the frequencies the human nervous system responds to before the conscious mind catches up, the sounds of the infected. Not screaming exactly. Something that had started as screaming and been processed into something else.
The assessment benchmark was one hour. Most candidates reached fifty minutes before the body's stress response overrode the mind's instruction to remain calm. The cognitive architecture of panic was well-documented and consistent: elevated cortisol, tachycardia, the complete reorientation of cognitive resources toward the single imperative of exit.
Alen floated in the dark water and listened to Raccoon City die.
He did not block out the sounds. He had tried blocking, in the first session, and found that suppression created a pressure — a building internal resistance that was itself a form of distress. Instead he had found something else. He let the sounds in. He let them be what they were — recordings of something that had happened to real people in a real city because a system had decided they were acceptable collateral — and he used them the way he used everything that was difficult: as information. As data. As the answer to the question why.
This is what it sounds like. This is what you are here to prevent. The sounds are not a threat. They are a briefing.
He floated. His heart rate, on the external monitors, read fifty-one beats per minute. His cortisol levels, measured through the dermal sensors on the tank's interior surface, remained in the baseline range. His brain activity, on the EEG feed, showed the focused, low-frequency pattern of deep concentration rather than the spiked, erratic signature of acute stress.
At four hours, the programme supervisor stopped the session manually. Not because Alen had failed. Because continuing past four hours with the current audio feed was outside the ethical parameters of the assessment protocol, and because the supervisor had been watching the monitors for three and a half hours and had developed a feeling he needed to write a separate report about.
He wrote the report. He filed it in the section of the programme documentation reserved for anomalies that required senior review.
He labelled the file the same word he had been using for this candidate since January.
Exceptional.
Then he sat back and looked at what he had written and thought that exceptional was not quite right, and that the word he actually wanted did not exist in any assessment rubric he had access to, and that this itself was probably the most significant data point he had generated in eighteen months of running this programme.
∗ ∗ ∗