Home

The Non-Native Articulatory Corpus (NNA Corpus) is an online database of speech samples from native and non-native speakers of different languages. Speakers are recorded reading aloud a standardized set of sentences using a microphone and an ultrasound probe, which provides information on the movement of the speaker’s tongue during articulation. The corpus is designed to be freely used by teachers, linguists, speech therapists and anyone else who is curious about foreign accents in different languages.

The intention of the NNA Corpus is to provide both audio and ultrasound video recordings as a free resource on the second languages spoken in and around George Mason University (GMU). Our long-term goal is to have recorded participants in each of the languages commonly learned as an additional language by English speakers, including those taught at Mason (e.g., Arabic, German, Spanish Chinese, Korean, etc.). The NNA Corpus began data collection with the recording of French non-native speakers in 2021.

The NNA Corpus will be presenting at the Acoustical Society of America on December 1, 2021 in Seattle, WA.

NNA Corpus utilizes a non-invasive method of viewing tissues and movement of the tongue. Ultrasound creates a sonogram of the tissues and organs within the body using high-energy sound waves without the participant’s discomfort. To read more about our project, check out our About page. Our materials and methods can be found on our Methodology and Resources pages.