The exercise was explained to her on a Tuesday morning in March, in the greenhouse, with the specific matter-of-fact tone that Mira used for things that were simply the next step rather than the culmination of anything.
"Respiratory compound," Mira said. "Slow administration. Ten-day course. Symptoms consistent with seasonal illness. Natural resolution within three days of last administration." She handed Lira the formulation notes. "You will design the dosing schedule, prepare the compound, administer it, monitor the progression, and achieve controlled resolution by day thirteen."
She indicated the subject of the exercise: a sheep in the pen beyond the greenhouse's south wall, a brown-faced lowland variety that had been moved there specifically for this purpose.
"Week one," Mira said. "Design. Week two begins administration."
She left.
Lira designed the schedule over four days, building the dosing curve with the care she brought to analytical work — considering the subject's weight, the compound's pharmacokinetic profile, the expected progression of symptoms, and the specific points at which she would need to assess whether to adjust. She built in three assessment intervals: day three, day six, and day nine. She wrote out each of the antidote and countermeasure protocols for the three most likely adverse reactions.
She showed the schedule to Mira, who read it without expression and made one modification: the day-three assessment criteria were too conservative. She expanded them.
The administration began on the eighth day of March. Twice daily, in the compound's delivery form, introduced to the subject's water. She noted symptoms on each visit — the respiration rate, the nasal evidence, the behavioral indicators of discomfort — and logged them in the journal with the precision she brought to pharmaceutical records.
By day four, the symptoms were consistent with her projected curve. By day six, she adjusted the dosing by twelve percent in response to an assessment showing slightly slower progression than expected. By day nine, the subject was displaying the target presentation, and she began the resolution protocol.
Day thirteen: full resolution. The subject was eating normally and showing no residual indicators.
She sat in the greenhouse on the afternoon of day thirteen, in the specific quiet of a completed task, and thought about what she had done.
The moral weight was present. She had made it present, deliberately, throughout the exercise — had not allowed herself the comfort of treating it as purely technical. On each of the thirteen days, during her observation visits, she had held the full awareness of what the exercise was preparing her for. This was the discipline she had built into her framework: the action and its meaning were to be held simultaneously, not sequentially. Not done first, understood later. Known throughout.
The subject was an animal and the exercise was controlled and its outcome was full resolution, and none of this was what the real application would be. She knew this. The exercise was not rehearsal for the specific circumstances; it was rehearsal for the specific internal management that the circumstances would require.
She wrote in her journal: the hardest part of this exercise was not the technical execution or the monitoring. The hardest part was holding the weight of what it was without letting it contaminate the precision of the work. The weight and the precision are not in opposition, but they require different parts of the mind, and those parts of the mind are not naturally comfortable neighbors.
She underlined the last sentence.
What she had discovered, over the thirteen days, was that the weight did not go away with familiarity. She had known, from the exercise's design, what she was doing and why. The knowing did not make the weight lighter; it made it cleaner. There was a difference. A light weight could be ignored. A clean weight could be carried without distortion.
She thought about Mira's words from months ago: a scar heals, but you always know it is there. She thought she understood this now in a way that understanding-as-concept had not given her.
She also thought about the reverse: the seventeen compounds she had been building antidote profiles for, the therapeutic applications she had mapped alongside the harmful ones, the full pharmacopoeia she carried in her head that included both sides of each possibility. She had not done this because Mira told her to. She had done it because the framework required it: every capability that was a weapon was also, potentially, a remedy, and the practitioner who only knew one direction was a more limited practitioner than the one who knew both. She would use this knowledge. She was already certain of it, though she didn't yet know in what specific circumstances.
She closed the journal. She went to check on the subject, who was standing in its pen in the March afternoon light, eating.
She gave it a handful of extra grain, which was not part of any protocol. It was something else — an acknowledgment, a small act of the kind that kept the weight honest. She stood in the pen for a moment with the animal's uncomplicated presence and the March afternoon around them and thought: I am not what I was, and I am not sorry.
She waited for the rest of the sentence to arrive, and it did, as it always did when she was honest: and I miss who I was.
Both things were true simultaneously. She had learned to let them both be true.