The New Yorker:

Americans tend to think that we’re a pretty homogeneous nation, in terms of our vocabulary. Yes, there are Southern drawls, and there’s Boston and Brooklyn and Appalachia and Minnesota, but the words themselves, we believe, are pretty much the same. But there are often significant regional differences, and these are beautifully explicated in the Dictionary of American Regional English, the six-volume study of America’s dialects, affectionately known as DARE.

There’s nothing terribly mysterious about the process of writing a dictionary. You figure out what you want to include, research it, and then write it up. But there are a lot of ways to do your research. If your subject is ancient Greek, you read everything there is, organize your words, and look at every example. If your subject is dialect, the problem is to figure out how people say things, and what people call things, in places that may not be that easy to get to. Several nineteenth-century projects, most notably the English Dialect Dictionary, managed this by sending surveys through the post and hoping to get useful answers back. A modern option is to remain in one’s office and telephone people around the country, the process used by the Atlas of North American English (which focusses almost entirely on pronunciation). The other obvious thing to do is go everywhere in person. In the late nineteenth century, Edmond Edmont rode by bicycle around France, and corners of Belgium and Switzerland, conducting seven hundred interviews to research what became the Atlas Linguistique de la France.

When DARE began its fieldwork, in 1965, its teams travelled in “Word Wagons”: campers outfitted with detailed surveys, recording equipment, and linguistics graduate students. The wagons travelled to more than a thousand communities, which were not always ready to welcome sixteen-hundred-question surveys, massive reel-to-reel tape recorders, or graduate students. But they persevered, completing almost three thousand interviews, with a total of 2.3 million answers that were keyboarded in a prominent early example of the value of computing in the humanities. Recordings of the interviews, and of a story called “Arthur the Rat,” which was written to elicit important pronunciation features, show the variation in speech sounds. This fieldwork, combined with extensive original textual research, formed the core of the dictionary, which was ultimately published from 1985 to 2013.

Go to link