Stimuli

The stimuli are composed of 60 French sentences (words per sentence: M = 10.5, SD = 2.24; total word count = 630) of varying types, including declaratives, imperatives, interrogatives, and direct quotes (see "Resources" section for target sentence list). Those sentences illustrate a wide range of phonemes (phonemes per sentence: M = 33.7, SD = 5.98), 46.6% of singleton consonants, 46.3% of vowels, and 7.1% of consonant clusters.

Procedure

Prior to recording, participants complete a detailed questionnaire in which they report their demographic information, language background, French language proficiency, and use (see "Resources" section). In a sound-attenuated booth, participants then proceed to the speech production experiment using ultrasound imaging to capture their tongue movements. The average recording time lasts approximately 30 min. 

            During the experiment, participants read the target sentences while both acoustic and articulatory data were simultaneously recorded using the computer program Articulate Assistant Advanced (Articulate Instruments Ltd 2012). The articulatory data includes the midsagittal tongue contours, captured with the Micro Ultrasound System and a head stabilization helmet (Spreafico et al. 2018). 

            Participants are seated in front of a laptop computer, wearing the Ultrafit head-stabilizing helmet (see illustration on this page and supplemental document in the Resources section) to ensure stabilization of the ultrasound probe under the participant's chin. Ultrasound images are obtained using the Micro Ultrasound System for Speech Research (Articulate Instruments), with a 64 element microconvex ultrasound transducer (2-4MHz, 20mm radius, 81.4 frames per sec, 92-degree max Field of View). During the entire procedure, ultrasound gel (Aquasonic 100, Parker Labs) is applied between the speaker's chin and the probe often times. To simultaneously capture the acoustic data, a lavalier microphone (Røde smartLav+) is also attached to the helmet (sampling rate: 22,050 Hz). Participants may browse through the stimuli once to briefly familiarize themselves with the sentences. They are also asked to read the sentences as naturally as possible. Finally, each participant is provided with a small bottle of water. At the beginning of the recording session, participants are asked to sip a small amount of water several times to obtain the palate trace (see ultrasound manual in "Resources" section for the complete setup).