Are circles an illusion, Euclid’s dream?—C=πd—a fiction we pretend is real in order to play marbles, build an arch, get a satellite into orbit? Does it matter? Who cares what gave a casting its shape once its mold is stripped away? Or if the scaffolding of statistics, population, blot-stain analysis, controlled trials, and all the rest did what we thought they did once the vaccine works? Or at least seems to.

Lots of wrong reasons for the right answer.

A dark pixel is a dot; three dots in a row is a line; a line that bends is a curve; a curve that bends back to its start is a circle; three dots in a circle is all an infant needs to recognize a face. And here our troubles begin: software straining to tell a smile from a grimace, when in reality no two cats are alike, no two lab rats, no two lab monkeys, no two human test subjects though we use one word—mammal—to gather features that are conserved across all. We make a vaccine for the group and hope it applies to the individual even if we know that a study’s average height of 5’4” might not match any one. We draw an evolutionary tree
when it should be a cloud: billions of clouds, ever changing, patterns that make up the cartoon we call us.