In what’s become a problem in their journalism courses, Chinese graduate students at Northeastern University have found that having English as their second language impacts their ability to correctly transcribe quotes.

Emerald Li is non-native English speaker currently in graduate school who was recently stressed out when she was unable to do a word-by-word transcription for a story.

“I attended a speech about religious issues a couple of weeks ago,” said Li. “I could not even make inferences of the proper nouns based on the context when I listened back to the recording. Finally, my professor commented my quotes fake and graded my paper [a] C.”                    

Similarly, Becky Chen, a first-year journalism graduate student recently struggled to find grammar mistakes in direct quotes for an in-class assignment.

“My professor pointed out that ‘cut his words’ makes no sense in grammar,” said Chen. “Instead, I should add ‘off’ with brackets so it works. However, when doing the transcription, I totally had no idea of the grammar going wrong. It’s difficult to listen to linking words in a speech.”

It’s a problem that some academics are trying to disentangle.

“One way of making sense of this problem is to think of the relationship between speaking and listening as an auditory feedback loop for people learning a second language,” said Christina Michaud, a senior lecturer in Boston University’s writing program.

“Typically, learners process sound that convince based on the things that they say,” said Michaud. “They will hear things probably the way they are more likely to say it. So this might cause them to transcribe quotes incorrectly.”

Marnie Reed, an associate professor of education at Boston University, says that the mismatch of the input and the learners’ own output leads to a lack of understanding of the target sound.

“For example, for regular count noun in plural and past tense regular verb endings, you need to add an additional syllable to the original syllables. But I’ve heard students try to compress ‘horses’ into just one syllable,” said Reed, adding that the mispronounced word is connected to an acoustic image that could affect learners’ understanding of “horses” in two syllables.

Then comes the question of how to break this feedback loop. Pronunciation professionals like Reed and Michaud propose a production-centered solution. In practical terms, when learners “finally get it right” and produce target-like sound repeatedly, they are able to form a mental model of what the word sounds like.

“Awareness is key,” said Michaud.

In their study ‘An Integrated Approach to Pronunciation,’ Reed and Michaud also introduce connected speech features — linked, reduced, deleted, altered and contracted sounds — as a premise for instruction.

“You have isolated vocabulary words, that is, how each of these isolated word pronounces, like ‘his,’ ” Reed said. “But without any instruction on what to listen for and how to listen, the person who starts by listening to a TV show may not aware that ‘h’ is completely deleted. So you end up with something like ‘Is his busy?’ where everything’s connected and the ‘h’ goes away.”

Although these challenges have strained Li , she sees it as just a bump along the way to her goal of becoming a successful journalist.

“Getting error-free grammar in news is hard. But the goal is worth the effort.” said Li.


Photo credit: Michaud and Reed.