Pronunciation Lab

Dedicated to sounds which are known problems for native English speakers, this app groups sounds, first by type (e.g. “Front rounded vowels” which may contain multiple IPA characters) and then by IPA character (e.g. "ø” of which there is necessarily only one). For each IPA character, there is a recording of a native speaker, providing a “base-truth” of a given practice text. Then the student would be able to record themselves and listen back to self-evaluate.

Unlike other shadowing techniques which focus on phrase pronunciation, this app would focus on sound pronunciation. This is helpful because it catches edge cases which are not currently being accommodated. We can imagine a student getting 95% of multiple phrases correct, but the remaining 5% are consistently the same sound. Such a student may be able to attain a very high level in a language while still retaining gaps in their ability to pronounce the sounds of a language. Television, music, and daily life in the target language are not enough if the inability to pronounce certain sounds has proved persistent. This has been my experience after spending several years in France and the experience of my roommate after her undergrad in Korea.

I’ve heard of the shift in language instruction toward accepting L2 accents. This is great news, and this app does not seek to regress that development. Instead, it is best considered as a precise intervention for students having hard-to-solve pronunciation problems that exceed the acceptable margins of an L2 accent and instead are truly difficult to understand. The value of this would be even more obvious for advanced speakers where their pronunciation difficulties have proved resilient to a great deal of speaking and listening.

If this is of research interest to you, we should talk about a collaboration—get in touch by email at james.mitofsky@gwu.edu.