Research Examines ‘SHtriking’ Sound Change in Raleigh

The way you say certain words likely sounds different from the way your grandparents said them.

The same disparity likely exists from town to town and based on gender, race and other demographic factors.  

NC State graduate student Eric Wilbanks studies how sounds change in speech. In one of his latest projects, the M.A. in linguistics candidate honed in on (str) retraction, or when /s/ is spoken more like /sh/ in (str) clusters (e.g., “SHtreet”).

Findings from his study, which examined the ongoing phenomenon in Raleigh, give insight into the patterned changes happening in North Carolina’s capitol city. His research poster, “SHtriking’ Change in Raleigh’s Speech: Acoustic Analysis of (str) Retraction,” won first place in the humanities category at the 2016 NC State Graduate Student Research Symposium.

Wilbanks also presented his findings at the 40th Annual Penn Linguistics Conference and plans to publish his work in the near future. We caught up with Wilbanks about his study and his plans for the future.

Q: What led you to pursue this research? How did you get interested in sound change?

A: The question of how, when and why languages change is actually one of the things that first got me into linguistics. What really interests me is the fact that, although each of us has our individual knowledge about how our language system works, we’re also at the same time small parts of a much larger language system that is dynamic and fluid. Although you can’t observe the trajectory of a language by examining just one person’s language, the patterns begin to get clearer as you zoom out to the language of the community as a whole. Learning about those macro-level patterns is what really excites me!

In terms of this specific research project, I chose to work with (str) retraction because most of the work on sound changes in the South (and sound changes in general) has been carried out on vowels. And really, when someone imagines a traditional Southern accent, the features that probably come to mind the quickest are the vowels (imagine what the vowels would sound like if a person with a strong Southern accent said “Dad”, for example). We know a lot about how vowels are changing, but less work has been done on consonants. Part of the inspiration for this project is to expand our understanding of the way sound changes with consonants progress.

Q: What are some examples of (str) retraction?

A: One of my favorite example words is “street.” Imagine that you’re first saying it as if it was spelled “SHtreet.” Now try and say it so it has the same “s” as in the word “say.” Odds are that in your normal speech the “s” in “street” is somewhere between the two. (Str) retraction is an ongoing sound change where many speakers of English are slowly shifting to say it more like “SHtreet.” You can see this in lots of other words such as “construction,” “restroom,” “stripes” etc.

Q: (Str) retraction appears to be a change popping up across the U.S. and abroad. What factors made Raleigh an interesting place to examine this phenomenon?

A: Raleigh’s a really interesting place to do linguistic research because of the unique demographic history of the area. In the 1950s and ‘60s, the Research Triangle Park opened and lots of tech companies such as IBM set up shop.

While these new businesses did hire a lot of workers from North Carolina, they also brought along a huge number of skilled workers from the North (hence the folk etymology of CARY as “Containment Area for Relocated Yankees”). These new workers brought different linguistic systems and ways of speaking along with them. The interesting question for linguists, then, is how such a dramatic shift in demographics like this influences the language of the region. What we observe is that in urban areas with lots of contact between people speaking different dialects, you end up with a process of “dialect leveling” in which people lose the more distinctive features of their dialects and come to a consensus on a new linguistic variety. (Str) retraction is just one of the features that’s defining the new character of Raleigh’s speech.

Q: Where did your data come from and what did you analyze?

A: My data came from a corpus of interviews Dr. Robin Dodsworth, NC State associate professor of English, has been collecting over the past seven to eight years. The interviews are typically about an hour long and are all with people born and raised in Raleigh. After the interviews are carried out, we can analyze the speech signal using quantitative acoustic methods. For (str) retraction, the main feature I’m looking at is the frequency of the sound; more “SH” like sounds have a lower frequency than “S” like sounds.

Q: What did you find? What do these findings mean? Why do they matter?

A: The main research finding is that (str) retraction is currently emerging within the community and is being led by young women. For many different sound changes in many different regions, young women have often been shown to be the most innovative members of the community, picking up new varieties and spreading language change, so it’s great that we’re observing similar patterning here. Importantly, the emergence of (str) retraction in the community lines up with the influx of non-Southerners for tech jobs in RTP, and this sound change patterns similarly to the changes to the vowel system.

Another important finding is that (str) retraction is strongest when the (str) occurs in the middle of the word, as in “restroom”, while the change is less advanced in words like “street” where it occurs at the beginning. This is exciting for linguists because often when we notice a new feature in a language and go to study it, by the time we look to see what people are actually doing, everyone has already adopted the new feature across the board! What we’re seeing in Raleigh, however, is the very beginning stages of the sound change. We’ve caught (str) retraction red-handed and can observe the change from its very beginning and test the predictions made by different theories and models of sound change.

Q: What type of future research is warranted?

A: One of the things I’m most excited to investigate in my next project is the role of the actual physical articulation of the sounds involved in this change. In our lab we use ultrasound imaging techniques to get real-time data on the position of the tongue as speakers produce sounds. We can use this data to look at the effects neighboring sounds have on the realization of (str). There have been a few papers that provide solid evidence for articulation playing a big role in starting this sound change. 

For example, when you’re producing the word “street” you don’t say each sound in isolation before moving on to the next one. Instead, your gestures for each sound overlap and you start getting ready for the next sound while you’re still making the current sound. Analyzing this coarticulation and overlap of gestures directly through ultrasound imaging will be a really exciting next step for this project.

Q: How have professors or your experiences at NC State helped prepare you for research like this?

A: One of the great parts of my department (and NC State in general) is the focus on giving students experience doing actual original research using the most innovative methods and theories. All of the courses in linguistics here have a culminating final research project, and professors push students to expand upon these research projects and develop them to be suitable for conference presentation and publication. Even at the undergraduate level, there’s a real environment which fosters original research; for example, in my junior year I started doing original research through an independent study mentored by a linguistics professor and was supported by an Undergraduate Research Grant.

Q: What’s next for you after grad school?

A: After I finish my M.A. at NC State this May, I’ll be going to the University of California at Berkeley’s Linguistics Department to pursue my Ph.D. in  August. After that the goal is a professorship in a research university, though I’ll likely be a post-doc for a year or two before I find the perfect position.