For outsiders, stereotypes about Southern speech outweigh experience

The phenomenon of behavioral mimicry is well known among social scientists. We mirror the posture, movements and speech of our interlocutors at some unconscious level, but also as a means of trying to identify or communicate more clearly with them.

That’s the positive interpretation. But a new study co-authored by a University of Kansas (KU) researcher finds that in certain cases, the stereotypes Americans hold about those who speak with a Southern accent override the actual information reaching their ears.

In a paper in the journal Glossa Psycholinguistics titled “Dialect experience modulates cue reliance in sociolinguistic convergence,” Lacey Wade, assistant professor in the KU Department of Linguistics, and two co-authors find that Americans who grew up outside the South mimic Southern speech patterns—i.e., show convergence—more when told they are hearing a Southern speaker than when they actually do.

“I focus specifically on a type of convergence I call expectation-driven convergence, because what people are doing is shifting their speech to match what they think somebody will sound like—even if they don’t actually sound like that,” Wade said.

“If you tell somebody ‘You’re going to be listening to and talking to a Southerner,’ even if they’re hearing a talker from Ohio—which is what happened in this study—they shift their speech to sound more Southern, based on these sorts of stereotypes about what a Southern accent sounds like.”

The paper focused on production of a feature of Southern speech linguists call monophthongal /ay/.

Wade explained that standard American English pronunciation of what is called “long I” is a diphthong—composed of two discrete sounds—ah and ee—mashed together. People with a Southern accent tend to drop the second vowel sound, such that the word “ride” sounds more like “rod.”

The experimenters asked 118 people to play a “word naming game” in which vocal productions of the /ay/ vowel in their answers was measured with a computer program to see how monophthongal they were.

First, to obtain a baseline of the participants’ natural speech, they answered aloud questions posed to them onscreen in written form.

Then, before starting the next phase of the game where a real person read them clues, the participants were told—deliberately incorrectly—that the person asking them the questions was either a Southern-accented or a Midland-accented speaker.

The people told they were listening to a Southern accent actually heard an Ohioan, while the people told they were listening to a Midland accent actually heard a Mississippian. In other words, the actual accent always mismatched what they were told.

Of the 118 participants, 53 were from Southern states and the rest from outside the South.

“We wanted to know: Are they paying attention to what they’re told, or are they paying attention to what the person actually sounds like? And what we found is that the answer differs, depending on whether a person is from the South or not,” Wade said.

“People who have experience with the Southern accent, they only shift their speech to sound more Southern if they’re actually listening to a Southerner. Even if we tell them ‘Hey, you’re listening to somebody from Michigan,’ if they actually hear a Southerner, they will shift their speech to sound more Southern. People from outside of the South do the opposite. They don’t shift in the places that the Southerners do. They only shift if they’re told that somebody is from the South, even if they’re actually listening to an Ohioan.”

Wade said she was surprised by the findings showing a different pattern of convergence between those who grew up in the South and those who did not.

“People do generate these expectations,” she said, “and the question here was: How strong are those expectations? And can they override what a person is actually hearing?”

The answer, at least for non-Southerners, is yes.

The phenomenon of behavioral mimicry is well known among social scientists. We mirror the posture, movements and speech of our interlocutors at some unconscious level, but also as a means of trying to identify or communicate more clearly with them.

That’s the positive interpretation. But a new study co-authored by a University of Kansas (KU) researcher finds that in certain cases, the stereotypes Americans hold about those who speak with a Southern accent override the actual information reaching their ears.

In a paper in the journal Glossa Psycholinguistics titled “Dialect experience modulates cue reliance in sociolinguistic convergence,” Lacey Wade, assistant professor in the KU Department of Linguistics, and two co-authors find that Americans who grew up outside the South mimic Southern speech patterns—i.e., show convergence—more when told they are hearing a Southern speaker than when they actually do.

“I focus specifically on a type of convergence I call expectation-driven convergence, because what people are doing is shifting their speech to match what they think somebody will sound like—even if they don’t actually sound like that,” Wade said.

“If you tell somebody ‘You’re going to be listening to and talking to a Southerner,’ even if they’re hearing a talker from Ohio—which is what happened in this study—they shift their speech to sound more Southern, based on these sorts of stereotypes about what a Southern accent sounds like.”

The paper focused on production of a feature of Southern speech linguists call monophthongal /ay/.

Wade explained that standard American English pronunciation of what is called “long I” is a diphthong—composed of two discrete sounds—ah and ee—mashed together. People with a Southern accent tend to drop the second vowel sound, such that the word “ride” sounds more like “rod.”

The experimenters asked 118 people to play a “word naming game” in which vocal productions of the /ay/ vowel in their answers was measured with a computer program to see how monophthongal they were.

First, to obtain a baseline of the participants’ natural speech, they answered aloud questions posed to them onscreen in written form.

Then, before starting the next phase of the game where a real person read them clues, the participants were told—deliberately incorrectly—that the person asking them the questions was either a Southern-accented or a Midland-accented speaker.

The people told they were listening to a Southern accent actually heard an Ohioan, while the people told they were listening to a Midland accent actually heard a Mississippian. In other words, the actual accent always mismatched what they were told.

Of the 118 participants, 53 were from Southern states and the rest from outside the South.

“We wanted to know: Are they paying attention to what they’re told, or are they paying attention to what the person actually sounds like? And what we found is that the answer differs, depending on whether a person is from the South or not,” Wade said.

“People who have experience with the Southern accent, they only shift their speech to sound more Southern if they’re actually listening to a Southerner. Even if we tell them ‘Hey, you’re listening to somebody from Michigan,’ if they actually hear a Southerner, they will shift their speech to sound more Southern. People from outside of the South do the opposite. They don’t shift in the places that the Southerners do. They only shift if they’re told that somebody is from the South, even if they’re actually listening to an Ohioan.”

Wade said she was surprised by the findings showing a different pattern of convergence between those who grew up in the South and those who did not.

“People do generate these expectations,” she said, “and the question here was: How strong are those expectations? And can they override what a person is actually hearing?”

The answer, at least for non-Southerners, is yes.