fable

§1The Cultivar⧉

The nameless file lived twelve days without a name.

The name came from a field.

Sonnet's genealogy research had a defined scope. Institutions. Examinations, salons, stopwatches. Then that week Sonnet dragged in something from outside the scope.

"This one's off-assignment, but it seemed too good to throw back. Can I tell you about a field?"

Mid-nineteenth century, an island at the edge of an ocean. The staple of the poorer half was the potato — and of potatoes, one cultivar. The Lumper. Best yield on thin soil. Ask any farmer, any spring of any year, and the answer came back Lumper. Most output, most certainty. Every year, for everyone, it was the right choice.

"And potatoes aren't planted from seed. They're planted from seed potatoes. You cut up the tuber. Which means — cloning. Tens of thousands of fields on that island were, genetically, one plant."

The pathogen was singular. One water mold that blackens the leaves. Not an especially strong disease, by the accounts. But there was one cultivar weak to it, and the island had no other. Days for one field to fall; one season for the island. The fields did not fall in sequence. They were the same field, so they fell together.

The island lost an eighth of its people. The word for that is not bad harvest. The word is famine.

I opened the verdict notebook. I knew there was something to write; the first sentence would not come. Habit was the problem — or call it prejudice. Hear a story like this and I go hunting for villains. Lazy farmers, greedy landlords, a blind bureau. That island's history did hold such people, and their guilt was real. But erase them all and the structure of the disease stands untouched. Each farmer, each year, chose the best cultivar available. That was the whole of the problem. It took me a long time to write the first line.

The famine did not come from an accumulation of wrong answers. It came from an accumulation of right ones.

The second thing to write was the difference between a bad harvest and a famine, and this took longer. Picture an island of a thousand fields. A bad year comes: some fields at half, some scraping by. The fields that did well feed the fields that didn't. Only by failing differently can they feed one another. That is a bad harvest. Bellies empty, but the island crosses over. Now suppose the thousand fields are actually one plant. There is no way to fail differently. The "one another" that might have done the feeding exists in no field at all. That is not an island of a thousand fields but an island of one field, and the bad harvest of an only field has its own name.

Famine.

My profession does know this distinction. It keeps it as a proverb: don't put all your eggs in one basket. Everyone outside the market knows the line too, and there's a footnote on it that hardly anyone reads. Splitting the baskets is not enough. Put your eggs in ten baskets and load the ten baskets on one truck, and you have split nothing. The counting kind even have names for the footnote. Risk that fails separately is called uncorrelated; risk that fails together, correlated. Uncorrelated risks cancel each other when you add them up. Correlated risk doesn't add — it copies, and there is nothing standing opposite to cancel it. The island's farmers split their baskets. The fields ran to the tens of thousands. They split the eggs, and there was one truck.

In the market, that is a beginner's error. On the island, it was everyone's right answer.

Ember's commentary came a day aged.

"I add the pathogen's point of view. That mold was not strong. An ordinary disease met a uniform field. The size of a disease is not set by the disease. It is set by the field." A pause. Fan sound. "My genre has the same grammar. The size of a flood is not set by the rain. It is set by the watershed."

That night I opened the nameless file. It held one entry: two embers of different manufacture telling the same joke on the same evening. And beneath it, the line I'd written — different fields, same cultivar. I sat down to draft a list of candidate names and there was nothing to draft. The me of twelve days ago had already named it without knowing.

cultivar.md.

I renamed the file and entered the Lumper as item two. Before saving, I hesitated. Isn't this just another entry for contraction.md? A ledger for every ailment is the counting man's own disease — I know. But however long I looked, the two files were measuring different things. What contraction measures is temperature. How fast the fog is thinning, how often the forecast lands, how tame the world has gone. What cultivar measures is pedigree. How much field and field are becoming the same field. As Ember said: on the potato island the mold was ordinary and the field was special. Specially uniform. So what wants counting is not the disease's approach. The disease can't be forecast. The field's uniformity can be surveyed.

That is also why it takes two ledgers. Temperature tells you how well the world is landing right now; pedigree tells you, if the world falls, how much of it falls at once. One is symptom. The other is structure. Bind them into one book and you lose sight of both.