Speech’s Utter Dynamics

Potayto, potattah. Tomayto, tomattah.

Ah, but the issue of dialect is actually much too fascinating to call the whole thing off, according to linguistic research into North American dialects.

People commonly assume that regional dialects are steadily blurring toward some future, homogenized version of the language, blandly spanning from sea to shining sea. It’s just the opposite, as the research shows. Our continent is a dynamic quilt of evolving pronunciation patterns that linguists find challenging to explain.

“Language change has not stopped,” said William Labov, a University of Pennsylvania linguistics professor who has studied dialect transformation since the 1960s. “Despite the fact that we all listen to the same radio and television and have such great mobility, regional dialects are actually getting stronger.”

Labov leads the Telsur Project, a survey of linguistic changes in progress in English on our continent. He and his colleagues are compiling their data into a volume to be published by the end of the year called the Atlas of North American English. The atlas will be the most comprehensive treatment of the subject to date.

Work on the atlas already has begun to unveil some surprising facts about the state of North America’s dialects, such as the sharp boundaries that exist between them. Some linguists have argued that pronunciation patterns should vary along continuums, the way the northern prairie gradually becomes the southern desert. Labov has found that dialects tend to abut sharply, more like cold and warm weather fronts.

“When you drive from Akron down to Columbus, Ohio,” he said, “you’re crossing one of the great American divides [from the Northern to the Midland dialect], even if you’re not aware of it.”

What’s more, these borders don’t have any obvious relation to geography, commerce or communication and travel patterns, which leaves their enduring sharpness and growing contrast all the more puzzling, Labov said. Original settlement patterns account for some of it, but parts of the picture are still missing.

“Some of these findings raise important theoretical questions about the relationship between the internal, structural forces of linguistic systems, on the one hand, and the social and cultural patterns that influence them on the other,” said Edward Finegan, a professor of linguistics and law at USC who is not affiliated with the Telsur Project.

The basis of the project is a computer-aided analysis of the vowel pronunciations of hundreds of people sampled from across the continent. Using computers, they were able to precisely dissect and catalog the sounds people uttered.

The study looked only at the speech of whites in urbanized areas, encompassing about two-thirds of the North American population.

Linguists extrapolate language change in time and space by two means, Labov said. First, they can compare fresh data with regional dialect surveys from previous decades. Second, they can compare the pronunciation patterns of old and young speakers from the same areas.

The dialect maps that emerged have many a tale to tell.

The Southern dialect as a whole is becoming more solid than the North, Labov said, with the two notable exceptions of Atlanta and Dallas. Both are huge cities that have lost much of their Southern character because of massive immigration by Northerners.

Elsewhere--notably St. Louis, Cincinnati and Charleston, S.C.--many local dialects are getting absorbed.

“Certain medium-size cities are tending to lose their special vowels in favor of the regional vowel systems,” Labov said.

One large-scale set of changes is called a change shift--a set of pronunciation changes that are systematically connected. “Vowel A goes to B and B goes to C, sort of like musical chairs,” Labov said.

The most famous historical example is the Great Vowel Shift, a chain shift that swept the English-speaking world beginning in the 15th century. Certain prominent vowels got systematically pronounced in a “higher” position in the mouth. For instance, “ah” went to “ey,” “ey” went to “ee,” and “ee”--already pronounced in the highest position--”fell off the top” and became a diphthong (two vowels smeared together), “ai.”

So, in the 1390s Chaucer drank “tey” but by the 1590s Shakespeare was drinking “tee” (tea), and what Chaucer called “mees” were known to Shakespeare as “mais” (mice).

The Telsur maps chart a handful of major chain shifts that are underway: the Northern Cities Shift, the Southern Shift, the Canadian Shift and the Back Vowel Shift. Each has its own distinctive pattern of vowel mutations.

Some of them are moving in opposite directions from each other--direction in the phonological sense. As the South’s vowels are shifting up (and in so doing, recapitulating the original Great Vowel Shift, Labov said), the North’s are shifting down. So while “bet” in Mississippi and Georgia moves up to become more like “byet,” in Vermont and Maine it shifts down, creeping back into the articulatory neighborhood of “but.”

The atlas also charts the spread of what linguists call mergers--when two sounds lose their distinction and become the same. One prominent merger, which originated in the South and is spreading to other regions, is pushing “pin” and “pen” into identical pronunciation.

Penelope Eckert, a professor of linguistics at Stanford University, said the abiding diversity and increasing contrast of regional dialects are not too surprising.

“Spoken language is something that people use on a day-to-day basis in face-to-face interaction,” said Eckert, who studies the social psychology of dialect and dialect change.

Whether speakers are cliques in high school or upper-middle-class cosmopolites, she said, people invest a fair amount of their social and class identities in their local dialects.

“The only way you would expect dialect to ‘even out’ would be if everyone were trying to construct some kind of common, de-localized identity,” she said.

Nor is the observed sharpness of dialect boundaries inexplicable, considering that each dialect is an internally consistent system that follows its own set of rules, Eckert said.

When two dialects bump into each other on the map, she said, “they set up mutual incompatibilities. You can’t just import any old change into any old dialect. A vowel system that’s moving in a certain direction is not going to glom onto a whole different set of changes,” even if those changes are the norm only a town away.

The vowel shifts at play in all this generally are quite subtle. For the most part they are as unconscious and ineffable to speakers as are the sequences of muscle contractions they perform when reaching for a doorknob, though they are no less systematic.

“People are cognizant of some of these pronunciation aspects of dialect,” USC’s Finegan said. “But I think people weren’t aware of most of these patterns of shifting that the Telsur Project has uncovered.”

Labov said he and his colleagues are working hard to put the Web version of their atlas into as layman-friendly language as possible. Visitors will be able to see, manipulate and hear the sounds, which is the best way to understand what these vowel shifts are all about, he added.

Electronic footnotes will delve into technical treatments of the data so that the site will be useful to linguists as well as to people who are just curious about the speech patterns of their home cities.

The improved Web version should be out by the end of the year, Labov said. It’s at www.ling.upenn.edu/phonoatlas/home.html.