A patient comes in with a swollen knee and a rash I do not like the look of. I read the film the way I was trained to read films, structure first, before she has said a word about how it started. I put my hands on the joint the way years at the bedside taught me to, feeling for warmth and fluid before I look at a single lab value. Underneath both, an older habit takes over, the one that wants the whole system rather than the joint or the film alone. None of these ways of looking asks the other two to step aside. They run at the same time, on the same patient, inside the same fifteen minutes.

None of these habits happened by accident. Each took years of actually being good at it before it became reflex. Reading a film first taught me to notice structure before anyone tells me what hurts. Time with pregnant patients taught me that a single body can, for months at a stretch, quietly run as two. By the time joints became my daily work, what I actually wanted to see was the place where a system stops agreeing with itself.

Last year I added a habit that nobody had to sign off on. I started writing the software myself: a platform for structuring radiographic reports the way I actually think through a film, not the way a template assumes every case will fit, rebuilt line by line until the code did automatically what my own head does after two decades of images. Nobody asked me to learn this, and practicing medicine did not require it. I built it because reading a film, feeling out a joint, and thinking through a whole system had never felt, in my own head, like three different people arguing over whose turn it was to speak. They were one habit, still going, that kept finding new material to work in.

There is a piece of biology that gave me better language for this than I ever managed on my own. Every woman is built this way at the cellular level, whether or not she ever notices. Early in development, each cell in a female embryo permanently silences one of its two X chromosomes, the maternal copy in some cells and the paternal copy in others, decided separately and close to at random, cell by cell. The classic proof is a calico cat’s coat: every patch of orange or black fur is the visible record of one founder cell’s decision, still legible years later. A woman’s body runs the same arrangement on the inside, invisibly, in tissue instead of fur: two slightly different genomic programs, coexisting in one body since before it had a face, neither one required to win.

It even has clinical teeth. Most genes on the silenced X stay quiet, but a few partly escape, among them TLR7, a gene that helps immune cells sense viral RNA. Researchers showed in 2018 that some immune cells in a woman’s body express TLR7 from both X chromosomes instead of one, and that extra dose is now one of the leading explanations for why lupus, in particular, affects women at a rate commonly cited around nine to one. My patients are, overwhelmingly, women, for a reason that traces down to exactly the mechanism I am describing about myself.

I still do not know whether my own X-inactivation happens to skew toward one parent’s copy or splits close to even. Most women never find out; there is rarely a clinical reason to look. But I no longer think of any of this, the exam-room instincts or the software, as a scattered set of interests that never resolved into one identity. A woman’s body was never running on a single self to begin with, down at the chromosome, and mine does not seem to have gotten the memo that the rest of me was supposed to specialize. I still read films now and then, I read joints for a living, and somewhere along the way I taught myself to read code too. None of it has ever needed the others to stand down.