## DataLab

A few days before we launched this website, in a dimly lit creative director’s office strewn with sketches and animal crackers, an argument erupted. We were actually there to argue over something else entirely — site logos — when the editor-in-chief’s head snapped up. “Oh, no!” he said.

He had spotted a sentence with the phrase “data are.” He was not happy about it.

Here’s a dramatic rendering of the debate that followed (heavily edited to make everyone sound smarter):

“It should be ‘data is.’ It just sounds wrong otherwise.”
“You can’t count data. It’s like rain. You would never say, ‘The rain are heavy today.’ It’s a mass noun.”
“Grammatically, though, it has to be ‘data are.’ You can’t argue with the grammar.”
“But ‘datum’ isn’t a word we ever use. So it makes no sense to use the plural when the singular doesn’t exist.”
“AP style says ‘data is.’ ”
“Actually, no, AP style says to refer to it as ‘data is’ when you’re referring to it collectively.”
“Whatever the rules are, ‘data are’ just sounds pretentious and abstract. It will alienate readers.”

You get the idea. We spent the next half an hour shouting and even picked up the argument the next day.

One of the things that makes the new FiveThirtyEight so exciting is that its staff members have a rich array of professional backgrounds. But that also means that we’ve had different grammatical structures ingrained into our writing (and reading). We finally decided on “data is” on the basis that our writing style would be more approachable if we avoid sentence structures that aren’t often used verbally.

On the whole, our readers — at least on Twitter — agreed (1,279 to 615 the last time we looked). We asked followers of @FiveThirtyEight to favorite the tweet with the structure they preferred. Arguments are good, and we intend to have plenty more of them.

Mona Chalabi is the lead writer for DataLab.

All Science

Filed under , ,