Four years ago, Yasaman Rafat became a mother and faced a dilemma that affects millions of the country’s non-native English-speaking immigrants.
“How much of my native language, Farsi, do I speak with my son compared to English? When we are outside, do we speak in Farsi or English?” asked the Modern Languages and Literatures professor, who emigrated from Iran to Canada in her teens.
“My husband, who moved to Canada at a younger age than me, speaks both Farsi and English differently from me. I was curious to see how our son’s languages will develop in comparison with ours.”
With an inherent love for languages – Rafat speaks Farsi, English and Spanish, with a knowledge of Serbian and French – her interest now extends beyond her immediate family to immigrant families across Canada.
Today, she is developing the Canadian Multilingual Speech Database, a site that collects and documents speech samples of multilingual immigrants who speak both in native languages and English.
“The database will allow scholars to understand how different native languages change when they come in contact with English and new varieties of spoken English may evolve over the years,” Rafat explained.
While most research in linguistics focuses on how a language is structured, Rafat focuses on how a language sounds. Specifically, she is exploring how second-language sounds are picked up by second-language learners, and how native-language sounds may change.
Using recordings of speakers in both English and their native language, she will analyze how, for example, the pronunciation of a specific word in Farsi changes over three generations of Iranian-Canadian immigrants.
The database is in its first phase, where speech samples are being collected and documented of the native/heritage languages and English from first-, second- and third-generation multilingual immigrants living in Ontario. Later in the project, data collection will expand to other regions of Canada.
Currently, there is an American Speech Accent Archive and an Australian National Database of Spoken Language, but no Canadian database representing both English and native/heritage languages of immigrants.
“Studies have shown within six weeks of learning a new language, the native language of immigrants gets affected,” Rafat said.
Older generations of Canadian immigrants fear the gradual loss of their native languages. Rafat hopes her database will help them preserve that cultural heritage.
Using the data, she will also explore how different groups of non-native English speakers pronounce specific English words.
“The accent is not perceived as a ‘good thing.’ If you can’t speak ‘proper’ English, therefore you can’t ‘think properly,’” Rafat said. “Foreign accents in English do not need to be ‘corrected.’”
By recording varieties of English spoken by immigrant communities in Canada, Rafat is also documenting an important part of Canada’s heritage. “Let us acknowledge this diversity and embrace it,” Rafat said.
She works with graduate students in Hispanic Studies and Linguistics who come from many different immigrant backgrounds. “They have been enthusiastically helping me in setting up the database and will be instrumental during the data collection.”
Rafat wants the database to be a resource for linguistics researchers all over the world.
“If a researcher in Australia wants to study how Spanish sounds change across three generations of bilinguals or multilinguals in Canada, they would be able to use this database and compare it to sound change in Australia,” she explained.
The database has numerous possibilities, including serving as a reference for actors who want to learn how to imitate different accents, software engineers working on accent-recognition technology or sociologists and historians interested in questions about migration, societies and language.