After moving to a new place, some people’s accents change readily while others stay more or less the same for the rest of their lives. McGill University linguist and Assistant Professor Morgan Sonderegger recently spearheaded a study that explores the science behind accent dynamics.

Some studies on accents have analyzed only one conversation with a subject, while others have spanned years of observation. Sonderegger, however, wanted to explore the relatively unknown realm of medium-term accent dynamics—how accents change over a timescale of months. 

While this is not the first study on the subject, Sonderegger said that it’s “certainly the most detailed.” 

Rather than using a lab setting, Sonderegger executed a “natural experiment.” He likened his research to evolutionary biologists studying birds on an isolated island—where participants  interacted only with each other in a removed, yet natural, setting.       The researchers used the British version of reality TV show Big Brother to examine how people’s speech changed over a three-month period. Big Brother follows a group of contestants who live in a house together and are continuously voted off the show by viewers until a winner is selected. Sonderegger was drawn to the United Kingdom because of the extreme accent diversity in a relatively small geographic area.

“We wanted to look at change in people’s accents over time […] and we realized it’s hard to do this because you have to actually be able to see people do this over time,” said Sonderegger. “This particular TV show was a good opportunity to do this.” 

This specific experiment was time-consuming because it was conducted in a natural setting and involved data from real speech, unlike uniform words and phrases that are analyzed in a lab setting. Focusing on the pronunciation of particular consonants and vowels, the researchers used recently developed software, like FAVE and AutoVOT, to analyze the change in sounds more efficiently. However, it still took more than a thousand hours of work and the help of 10 undergraduate research assistants to transcribe speech and write the programs to analyze the data.

In an isolated setting like Big Brother, where participants interact only with each other for months at a time, their accents were projected to change over the longer term. That wasn’t the case. 

“People don’t actually come to sound like each other over three months,” Sonderegger said, though he noted that a few people with closer relationships did, such as a pair who dated during the show. 

The team also found that “[t]here were big daily fluctuations in the exact way people speak,” according to Sonderegger.

Lastly, they deduced that there is remarkable variability in how accents change over time, and that change is dependent on the person. Like previous studies, it demonstrates the complexity of accent change over time. 

There’s some evidence for more culturally significant vowels being less susceptible to change. For instance, the way that some Brits pronounce “but” is extremely dependent on region and has been that way for a long time.

Overall, the question of why there are “changers” and “non-changers” when it comes to accents can be attributed to differences between people—researchers just don’t really know what those differences are yet. In fact, very little is known about which factors affect pronunciation over time, which is a good foundation for future work in the field.

This research relates to differences in language learning ability and cognition among individuals in general, and the findings fall nicely in between previous ones of short-term and long-term accent dynamics. 

“It ties [past research] together nicely, and that piece has been missing,” Sonderegger said.