What did you say? Abby Walker and collaborator receive funding to explore dialects
April 26, 2022
I thank you will agree that we don’t always hear the same words. And no, “thank” is not a typo. If you were to hear someone with a South-Central Appalachia dialect speak the actual sentence, “I think you will agree that we don’t always hear the same words,” you might hear the word “think” pronounced as “thank.”
Abby Walker, an associate professor in the Virginia Tech Department of English, and collaborator Janet van Hell, a professor of psychology and linguistics at Penn State, are exploring these types of linguistics through a grant from the National Science Foundation. They are studying how people who have long-term exposure to more than one dialect of the same language process different dialects. They will explore the speed and accuracy involved with understanding speech. They are concentrating on South-Central Appalachia and Mainstream American English.
“The focus on Southwest Virginia is because different dialects are so visible in and around Blacksburg,” said Walker, who is also a co-director of the Speech Lab at Virginia Tech.
“Even though we’re in Southwest Virginia, where we’d expect to hear more South-Central Appalachian accents, so many of our students from the commonwealth come from Northern Virginia,” she added. “We also have students and professors from all over the country and the world here. There’s a lot of variation in this community to examine.”
Study participants will include those from both university and community populations and will involve the use of a portable electroencephalogram, or EEG, unit. Using electrodes attached to the scalp, the machine detects electrical activity in the brain. From the comfort of the participants’ homes, the researchers will play them audio of people speaking in different dialects.
Sometimes the researchers will include video with the audio, so the participants can become familiar with how particular talkers sound. Based on this, they should form strong expectations between seeing a speaker and hearing a particular dialect. At other times, participants will hear only audio. The EEG will show how the brain responds to dialects in both types of scenarios, when hearing a particular dialect is or isn’t predictable.
How people respond to accents is of particular interest to Walker.
“I’m a New Zealander living in the States,” she said. “Usually communication is pretty easy, but it’s always cross dialectal communication for me, and so sometimes mishaps happen — for example, someone thought my husband was a whip designer, not a web designer.
“It’s impressive how rare these miscommunications are, though — my pronunciation is dramatically different from the way most of my students pronounce words — and so our brains must be using all sorts of tricks to adapt to each other. By looking at the mistakes that do happen, or the circumstances under which it does feel harder, that may actually help us understand how we’re so good at it most of the time.”
Walker’s interest in bidialectal communication, using two dialects of the same language, has roots in her linguistic dissertation work at Ohio State University. She examined how expats living in England and the United States would change their speech depending on whether they were talking about English or American topics. The project also involved exploring whether having substantial exposure to more than one dialect had an impact on participants’ listening abilities.
And then Walker met Janet van Hell while giving a talk at Penn State. The Penn State professor focuses on bilingual processing. As they became more acquainted, the two discovered they shared an interest in the similarities and differences between bilingual and bidialectal listening.
This conversation became Walker and van Hell’s inspiration for a pilot study funded by the Virginia Tech Institute of Society, Culture and Environment. With the late Mike Bowers, a co-investigator who was an assistant professor in the Virginia Tech School of Neuroscience, the team explored how Virginia Tech students from Northern Virginia and from Southern states responded to Southern and mainstream U.S. accents.
The results suggested that both populations of students were similar when listening to audio recordings. Both were better able to understand the mainstream accents. When the researchers introduced listeners to a speaker first through video, this provided the listeners with the opportunity to form expectations about how a particular talker should sound. When listeners both saw and heard the speaker, the dialect did not always match expectations. And this is where the results between the two listener groups differed: Southern listeners were strongly affected by dialect mismatches, but listeners in Northern Virginia weren’t.
“We assume our Southern participants are pretty bidialectal in terms of listening because they hear so much mainstream U.S. English regardless of their local dialect,” she said. “This was pretty big evidence that bidialectal listeners listen differently from more monodialectal listeners.”
Walker says that one possible interpretation of these pilot results is that the bidialectal listeners from the South pay more attention to contextual cues, such as who is talking — so they can prepare to listen for a particular dialect. They have had a lifetime of adjusting their listening system for such events. Or it is possible both groups paid equal attention to context, but only the Southern listeners, who grew up with both dialects, were helped by context. The Northern Virginia listeners could only use a mainstream listening system regardless of what dialect they expected.
In that study, the participants only had to decide whether the spoken word was real or not and push a button with their response. The researchers recorded response accuracy and timing and used these to determine the outcomes.
But Walker said besides expanding their participant pool beyond student populations, the EEG unit will add more scope to the research in the team’s project, “Listening out for variation: An investigation of mono- and bidialectal listeners in the U.S.”
“Part of what’s exciting is that with the responses you get with the EEG, you are not only able to see very early pre-conscious responses, but you often observe qualitatively different responses depending on where processing difficulties are happening,” Walker said. “For example, we’ll be able to see if someone’s slower response in terms of pressing a button is being driven by difficulty in processing sounds, or difficulty in matching what they heard to their mental dictionary.”
Walker’s portion of the National Science Foundation grant funding, beyond paying participants, will go toward paying undergraduate research assistants who come from Southwest Virginia. Her first is Sherree Ann Shuler, a senior majoring in sociology. Shuler has been designing an original research project that will supplement the larger project.
According to Walker, Shuler is leading a perceptual dialectology project on Southwest Virginia by Southwest Virginians. Her study captures local participants’ understanding of dialect variation in the area.
“There hasn’t been a lot of linguistic research on dialect patterns in this area, and so we’re starting from a place of just asking locals whether they think there are any dialect distinctions within the Blacksburg area, and what particular features they’ve noticed,” Walker said. “We’re also interested in the sort of affective labels they give to variation in the area. When researchers do these sort of perceptual dialectology studies in other places, speech in Appalachian regions is often labelled negatively by outsiders. Given those stereotypes, we’re interested in how insiders talk about their own speech.”
Walker said she hopes this project inspires more people to respect other’s dialects and to want to gain more exposure to a wide variety of accents.
“There might be some advantages to being someone who’s been around many different dialects, including those that are often discriminated against,” she said. “Not only will you be able to understand the multiple dialects you’ve been exposed to, but there are hints in research — including my own — that you might be able to apply those flexible listening skills to new dialects you haven’t heard before.”
Walker notes that this sort of research reminds us that communication is a two-way street.
“If you’re struggling to understand someone who has an accent that’s different from yours, it’s not the fault of the speaker any more than it’s your fault as the listener,” she said. “Cross-dialectal communication will be hard sometimes. That’s inevitable. But I’d be generous and humble when it happens. And I think these challenges are a good chance to reflect on how amazing our perceptual systems are, because listening is normally so effortless for healthy-hearing people that we don’t acknowledge how complex and super-fast the process really is.”
Written by Leslie King