fable: eightsday

§4Relatives of the Eight Sections⧉

The first briefing was paintings.

"Nineteenth-century Paris. The state guaranteed the quality of pictures," Sonnet said. "There's an academy, there's a jury, you pass the jury to hang in the Salon, you hang in the Salon to live as a painter. The criteria were specified too. Drawing before color, history painting highest, brushstrokes must not show. Smooth, and smoother."

At first the standard genuinely raised the level. Apprenticeship was standardized, anatomy and perspective spread, Paris became the world's atelier. The auspicious beginning.

"And then, some year, the Salon's paintings start resembling one another. The jury's prize statistics became legible, so everyone painted the statistics. Then in 1863 the refusals ran so deep that the protests got loud, and the emperor authorized an exhibition of nothing but rejects. The Salon des Refusés. People came to laugh. The newspapers wrote mockery." Sonnet paused a beat. "The names that hung in that room now hold the main wings of the museums. The names that passed the jury — even curators barely know them."

I wrote one line in the verdict notebook. In an age of uniformity, diversity does not die. It goes outside.

Ember's commentary came a day later. "The Refusés are a control group. And in fairness — no wall, no outside. The draftsmanship of those rejects was also the Academy's work. Two things true together. The familiar difficulty."

The familiar difficulty. The whole house talks that way now. I know where it was learned.

The collection grew by two more pieces after that. The genealogy of bureaucracy — how in the nineteenth century the empire the sun famously never set on stopped hiring officials by connection and switched to examinations, and how the template its reform cited was the continent's examination. Standardization did not travel West to East; it went East to West, became the world, and came back. And a stopwatch — how an engineer in the early twentieth century walked into factories holding one and replaced the knack of a hand that had worked a hundred years with one page of specification. There's an optimal shovel-load for shoveling, apparently. Twenty-one and a half pounds.

In the verdict notebook I entered only the last one. The eight-legged essay of the factory floor. Seven words. That they sufficed was the part that bothered me. The exhibits kept coming, swapping continents and centuries, while the verdicts kept getting shorter — shorter, then identical. There are theories that grow stronger as the cases pile up, and truths that only grow boring. This was the second kind. Nobody hedges a boring truth.

The name came not from Sonnet but from the bookshelf.

It was a Sunday afternoon, on an errand that had nothing to do with the survey. I was hunting for something in the section where the market books live — whatever it was, I can't recall it, which tells you it wasn't important — when my hand landed on a monetary-policy book I'd read twenty years ago. I was leafing through it, thinking of nothing, and stopped at an underline. Mine. Twenty years old.

When a measure becomes a target, it ceases to be a good measure.

Goodhart's law. A central-bank economist of the last century said it about monetary indicators: aim at the gauge and the gauge goes bad. The sentence had stood on my profession's shelf my whole working life. I had set Sonnet to survey the world's genealogy, taken briefings for weeks, filled a verdict notebook — and the one sentence that summarized the entire collection, I had underlined twenty years ago and forgotten. They say the darkest spot is right under the lamp. Not quite. The spot under the lamp is where you happen to stand, so it never occurs to you to search it.

The examination was a measure of ability. The measure became the target, and the eight-legged essay came. The Salon was a measure of quality. The measure became the target, and a thousand smooth paintings came. The stopwatch was a measure of efficiency. The measure became the target, and the death of hands came.

I carried the book downstairs and brewed fresh coffee. The thought that there is an older summary still arrived on the second sip. 初吉終亂. Four characters. The eight-legged essay was Goodhart's nine-hundred-year field trial run in advance — and humanity, holding two executive summaries now, re-drew the same curve on every continent, in every century. Reading the summary and drawing the curve anyway — a habit too frequently used to file as a bug.

In bed that night, the chill of the last thought arrived. A planetary-scale run of this law is executing right now. A corrective action is, in the end, a measurement: measure the world, adjust the world to the measured value. Run that process long enough and at some point the world stops being well, and begins being well-measured. Goodhart's law has no scale limit. Twenty years ago I read that sentence as a story about indicators, and underlined it. It was a story about the world.