The Sentence That Stayed

The sentence first appeared as a qualification, in a footnote, and it said that the effect had been observed in one setting and might not hold elsewhere. Torres wrote it because it was true and because she expected somebody to ask.

The first paper to quote it kept the qualification and put it in brackets. The second kept the brackets and dropped the setting. By the fourth, the sentence had become a statement about the effect in general, and it carried her name.

Nobody along the way did anything dishonest. Each step was a small shortening of the step before it, made by somebody who had read the paper in front of them rather than the one at the beginning.

Lindqvist found the chain eight years later, when a reviewer asked him for the original setting and the citation led him back through all four papers. He wrote it up as a note about citation and not as a complaint about anybody.

What the note says, in the end, is unglamorous and worth keeping: a qualification survives exactly as long as somebody keeps reading the thing it was written into. After that it is a sentence, and a sentence travels well.

Torres, who read the note before it was published, made one addition to it. She had expected the question, she said, and the question had come — twice, in the first two years, from readers who had gone back to the footnote. What she had not expected was that the question would stop coming.

The two of them disagreed about the remedy and said so in the note itself. Lindqvist would put the setting into the sentence, so that shortening it becomes visibly wrong. Torres thought that this only moves the problem one step along, because a sentence with a setting in it is a longer sentence, and the next reader shortens what is longest.