I have run this particular model many times in the past months. Not the catastrophic scenarios, the ones I constructed with such detailed dedication in my early actuarial panic about this relationship, the ones about social friction percentages and interspecies compatibility indices and the specific and numerous ways in which I was likely to ruin something I had no right to possess. Those models have been deprecated. I ran them into the ground and found their assumptions flawed and retired them with the same thoroughness I applied to any model that consistently failed to predict observed outcomes.
The model I run now is different. It takes all of that data and it produces, with a consistency and a confidence interval that would satisfy even the most rigorous peer review, a single output.
She is mine, and I am hers, and this is the correct and final model, and I am not running a revised version.
We stop at the drum circle pavilion, where the evening ceremony is building toward its opening sequence, the lead drummer laying down the foundational rhythm that the others will build upon, and the quality of the light has shifted further into gold and the air has cooled slightly and Livia has finished her skewer and is watching the drummers with the, interested expression she has when she is learning something new. I fold myself down to sit on the broad stone bench along the pavilion's edge, which brings me closer to her height, and she sits beside me and immediately leans into my side, and I put my arm around her, and she fits.
"This is good," she says. She means the drums, and the festival, and possibly the skewer, but she says it in the tone she uses when she means something larger and less specific, when she is doing what I have come to recognise as her version of the chest warmth I cannot quantify, translating it into the smallest and most practical available container.
"Yes," I say. I look at the side of her face, at the glasses and the black hair and the small, loose smile she wears when she doesn't know she's being observed. "Livia."
She turns to look at me.
I have thought, over these months, about whether there is a correct way to say this, a formulation that is precise enough and true enough and not so enormous that it becomes clumsy. I have drafted versions of it in my head with the same careful attention I give complex models, testing the load-bearing assumptions, checking for structural weaknesses. I have found that every draft fails in approximately the same place, which is the place where the language runs out of resolution for what it is trying to represent.
I have concluded that this is acceptable. Some quantities resist tidy expression. The appropriate response to a quantity that resists tidy expression is not to pretend it is smaller than it is.
"I intend to spend the rest of my life making sure your spreadsheets balance," I say. Her mouth quirks. I continue, because I started and the model commits. "Your professional spreadsheets, your personal ones, the informal ones you run in your head when you are calculating whether something is worth the risk." I look at her face. "I am an unusual variable in any standard model. The error margins on an interspecies relationship are non-trivial. I ran those numbers at considerable length before I understood that I was optimising for the wrong outcome." I pause. "I was trying to calculate whether it was safe. I should have been calculating whether it was worth it. Those are different questions, and they produce different answers."
She is very still beside me. Her eyes are doing the thing they do when she is processing something important, not the quick, efficient processing she applies to financial data, but the slower, deeper kind, the kind that moves through her face in small increments.
"And the answer?" she says.
"The current model," I say, "has the highest confidence interval of any model I have ever run." I hold her gaze. "You are worth every non-standard variable. You are worth every recalibration. I am not going to file myself down or compress my shoulders or apologise for the space I occupy, because you have been abundantly clear about your preference, and I trust your data." I reach over and find her hand, the one with the bracelet, and close my fingers around it with the careful, specific gentleness I have been calibrating for months, the force that holds without constraining, the grip that says I am here withoutsaying I am not letting go, though both are true. "I love you. That is the output. I have checked my work."
The drums behind us build toward their first full statement, the overlapping rhythms finding each other and locking in, and the pavilion fills with the resonance of it, and Livia is looking at me with her whole face open in the way that still, every single time, short-circuits my available processing.
She lifts her free hand and presses it flat against my chest, directly over my sternum, where the warmth lives that I cannot quantify.
"I love you too," she says. "I didn't need to run the model. I already knew."
The drums are very loud now and the gold light is everywhere and around us the festival continues its ancient, enormous life, and I am sitting on a stone bench with my arm around a small human accountant who is wearing Orc courtship gold on her wrist and looking at me like I am not a problem to be solved or a risk to be managed or a variable to be controlled, but simply, entirely, hers.
I tighten my arm around her.
She leans in.
The drums play on.
Three weeks later…
I have spent fourteen hours running probability models on the optimal method of proposal delivery, and I have concluded, with a degree of scientific certainty that would satisfy even the most rigorous peer review, that there is no optimal method, and I am going to embed the ring in a raspberry-filled donut and hope for the best.
The morning is bright and cool and entirely unscheduled, which is itself a small revolution in my operating system, the concept of spontaneity having been recently introduced into my regular processes by Livia with the firm insistence thatI am allowed to do things simply because I want to, not because they have been planned in advance and optimized for maximum efficiency. I have taken this instruction seriously. I have practiced. This morning I woke her at seven with coffee and the information that we were leaving the apartment in twenty minutes, destination undisclosed, and she looked at me with sleepy bewilderment and then genuine delight and said "Who are you and what have you done with my actuary," and I did not answer because I was too busy experiencing the chest warmth.
The bakery is small and human-scaled and Livia's favorite, the one she walks to on Saturday mornings when she wants to sit with a book and something sweet and be quietly alone in public. I have been here exactly four times with her and seven times without her, the latter visits occurring during my lunch breaks when I needed to confirm the structural integrity of my plan with the owner, a sharp-eyed woman named Miriam who listened to my entire proposal with her arms crossed and then said "You want me to hide a diamond ring in a donut" in a tone that suggested I had just described an actuarial model with a fundamental algebraic error.
"Yes," I said.
She squinted at me. "And you think she won't break a tooth."
"The ring will be positioned in the center of the filling, embedded in a manner that ensures discovery before mastication," I said. "I have brought dimensional specifications."
I produced the folded paper from my jacket pocket. Miriam unfolded it, looked at the extremely precise diagram I had drafted showing the exact depth and positioning of the ring relative to the donut's structural layers, and made a sound that I have learned to interpret as human amusement when it is being suppressed for professional dignity.
"You measured the donut," she said.
"I measured several donuts," I clarified. "For consistency."
