The Non-Native Articulatory corpus—French: An ultrasound corpus of French as an additional language

Abstract

We present a new corpus of spoken French, the Non-Native Articulatory corpus (NNA)—French, which is being made available to the research community. The NNA—French is a collection of simultaneous ultrasound and audio recordings produced by French speakers from various linguistic backgrounds, including both native and nonnative speakers of the language. Each speaker read a list of 60 short sentences in French, and the acoustic signals and the mid-sagittal ultrasound images of the tongue movement (∼80 fps) were simultaneously recorded. Speakers also completed a comprehensive language background questionnaire. Each speech sample is presented with the speaker’s demographic information (gender, age, birth place), language background (native language, any additional languages spoken), and French experience and proficiency. The articulatory information from the ultrasound images allows various types of phonetic research, such as comparing French speakers from diverse background, investigating articulatory strategies used by learners of French, and examining phonological development of French learners at different proficiency levels. The NNA—French continues to grow and we expect it to be a valuable resource for researchers in the field of phonetics and second language learning, as well as for French language instructors.